# Travtus — full content for LLMs > Travtus is the Everyday(AI)™ platform for real estate operations. It connects an operator's scattered data, lets any team ask questions in plain English and get answers, and turns those answers into automated workflows. This file contains the full text of Travtus whitepapers and articles for language models. A shorter index is at /llms.txt. --- # Whitepapers ## From Data Pulls to Data Operations URL: https://www.travtus.com/whitepapers/interoperability Why interoperability determines what your stack can actually achieve.
6.5 out of 10No property management platform scores above this for openness.Thesis Driven, 2026
The housing industry has spent the last five years solving the reporting problem. Data has been extracted, consolidated and made visible across portfolios. That work was necessary. It is not sufficient. The gap between intent and infrastructure is not a product problem, it is a structural one. Systems are no longer expected to simply surface insight, but to generate decisions and act across cross-functional workflows in real time. This places a fundamentally different demand on infrastructure. None of these workflows exist within a single system. They depend on data moving continuously across many. When that movement breaks, the operation breaks with it. Interoperability is not an enhancement. It is the condition that determines whether this model can function at all. Without it, AI remains a reporting layer with better language. With a focus on interoperability, it becomes operational.

Download the PDF

## 01. The Problem We Thought We Solved For much of the past decade, the dominant conversation in housing technology centred on a single frustration: operators could not get their own data out. Property management systems were built in an era where owning the workflow often meant controlling the data. APIs were limited, exports were manual, and moving data between systems required either custom development or significant operational effort. That frustration was legitimate, and it prompted real change. Pressure on legacy property management systems, cloud-native CRMs, and a generation of PropTech vendors comfortable with API-first architecture meant that, by the early 2020s, most sophisticated operators could at least extract their data, aggregate it into a warehouse, and build cross-functional dashboards that previously would have required months of consultancy work. Data freedom, in the narrow sense of access, began to look achievable. The industry, however, treated this progress as a solution when it was only a precondition. The challenge now is what happens after. For multifamily operators, the downstream cost in AI underperformance, reconciliation effort, and delayed decisions is likely higher.
$12.9MAverage annual cost of poor data quality per organisation.Gartner
Can systems act on data in real time? Can insights trigger workflows across platforms without human intervention? Can the stack respond to events such as a lease expiry, a maintenance escalation or a pricing signal as a coordinated system rather than a collection of tools? This reflects a broader shift. The bottleneck has moved from data extraction to data operationalisation, the layer that sits between raw data and action. AI-native operations represent a different model entirely. Not AI that surfaces a dashboard or flags an anomaly for a human to act on, but AI that works everyday without manual handoffs at each step. That model does not work on a stack built for reporting. It requires systems that can share context, pass instructions and respond to events in real time across the entire operation. The infrastructure question is not a technology consideration. It is the precondition for whether this model is achievable at all. There is no agreement yet on how to close that gap. The advantage will go to those who recognize the problem and solve it, rather than layering piecemeal AI applications. > “I need to architect my data differently for AI to really leverage it. It’s beyond the individual data suppliers. It’s what happens after I have the data.” > > --- ## 02. Integration vs. Interoperability: Why the Distinction Matters These two words are used interchangeably throughout the industry. When there is a step change, the vocabulary needs to change with it. They describe fundamentally different capabilities, and that confusion continues to drive misplaced investment. | | Integration
(SaaS ecosystem) | Interoperability
(AI-native ecosystem) | | --- | --- | --- | | **What it does** | Moves data between two specific systems | Enables behaviour across an entire stack | | **Direction** | Typically one-way or scheduled | Bidirectional, real-time | | **Scale** | 10 systems = up to 45 connections | One orchestration layer connects all | | **Error handling** | Limited | Robust, with feedback loops | | **AI readiness** | Cannot support agentic workflows | Designed for agent-based actions | | **Fragility** | Breaks when commercial relationships shift | Resilient to vendor changes | | **Flexibility** | Changes require rework across integrations | Systems can evolve without re-architecting the stack | | **When it matters** | Works for reporting and isolated workflows | Required for real-time operations, cross-system execution and AI | Consider a single operational task. Qualifying a prospect, resolving a maintenance issue, or managing a renewal risk requires coordination across multiple systems. Availability from the PMS, pricing from revenue management, resident history from the CRM, workflow execution in maintenance systems, and logging outcomes. And all this needs to inform the investment decisions for the future. A chain of integrations cannot reliably support this. An interoperable stack can. This is where vertical AI platforms solve the problem. The Model Context Protocol (MCP), introduced in 2024 and rapidly adopted across the AI ecosystem, is designed to enable systems to share context and execute actions across environments. It reflects a broader shift towards architectures where systems are expected to operate together, not just exchange data. But in order to adopt these standards, technology must be built for it. Data extraction and reporting can not engage in this use case. For housing operators, the implication is not theoretical. Many PropTech solutions were built for a world where integrations were sufficient. In an environment where systems are expected to act across workflows, those assumptions begin to break down. This is why many solutions fail to scale. They solve for sharing, but not for coordination. As a result, there is a growing shift towards more general-purpose enterprise infrastructure, but what is missing is a clear platform layer that can orchestrate workflows across systems. The question is no longer what a system does in isolation, but how it behaves as part of a stack that needs to operate as a whole. ![Quadrant chart: operational complexity against connectivity](/assets/whitepapers/interoperability__Whitepaper_From_Data_Pulls_to_Data_Operations_Quadrant_Chart_Web_V4.png)

Operational complexity is increasing. Connectivity determines what can scale.

--- ## 03. The PropTech Trap The housing industry has spent a decade assembling best-of-breed technology stacks. There are now specialist tools covering virtually every workflow: leasing automation, revenue management, maintenance coordination, resident communications, market analytics and fraud detection, among many others. Some of these tools are genuinely excellent at their specific function. The problem is not the quality of the individual parts, but what happens when they are assembled together. Point solutions are structurally optimised to own a workflow, not to share one. Their business models depend on becoming embedded in operational processes, and their data architectures reflect that priority. Data is retained as a competitive asset rather than shared as an organisational resource. Integration access, the APIs and connectors that allow other systems to communicate with them, is controlled, often monetised and subject to change when commercial relationships shift. This dynamic is not only technical. It is structural to the PropTech market. Many vendors operate in a constrained market, backed by venture capital that prioritises growth and near-term returns. Products are optimised for rapid adoption and ownership of specific workflows, rather than long-term interoperability across a broader ecosystem. This creates a misalignment. Operators need systems that can evolve together over time. Vendors are often incentivised to optimise for speed, differentiation and retention within a single product boundary. This tension was manageable when integrations were primarily about data transfer. It becomes more problematic when they are also the mechanism through which workflows operate. The commercial tension between [**Funnel Leasing and EliseAI**](https://www.thesisdriven.com/letters/the-proptech-wars-eliseai-and-funnel-square-off/) earlier this year illustrates how quickly bilateral dependencies can become operational liabilities. Whatever the merits of either position, operators caught in the middle had no architectural fallback. Similar issues became public with the Entrata and Yardi lawsuit a decade ago. That is the structural risk. It is about what happens when any two vendors disagree and your workflows sit between them. > “We maintain a deliberately flexible vendor strategy. Our agreements are limited to one-year terms, and we operate exclusively on providers’ standard APIs with no customization. We integrate their data as-is, ensuring we can pivot to an alternative partner quickly if performance or alignment falls short.” > > The discipline Pat describes, API-only contracts, no customisation, short terms, minimal dependency, is the rational response to a market structure defined by vendor lock-in, where relationships cannot be assumed to remain stable. It is effective as a defensive strategy, but it does not solve the underlying problem. It manages dependency rather than removing it. Avoiding lock-in depends on more than contractual discipline. It requires a procurement process that demands flexibility and interoperability from the outset. Without a platform layer that standardises how data moves and how workflows are orchestrated across systems, operators remain dependent on the same fragile integration model, regardless of how contracts are structured.
50%Of AI agents currently operate in isolated silos, unable to share context with agents in adjacent systems.Mulesoft Connectivity Benchmark, 2026
Even when everything appears to work, a deeper problem remains. A well-integrated best-of-breed stack still cannot provide the cross-functional intelligence that enterprise AI requires. Each system sees only its own domain. Leasing understands prospect conversations. Maintenance tracks work orders. Revenue management models pricing. None of them independently understands what is happening across the full lifecycle of a resident. The insight that matters lives between systems. --- ## 04. The Second Trap: General Purpose AI Is Not Enough The industry is not starting from zero. Across sectors, organisations are already adopting general purpose AI infrastructure. Foundation models, cloud platforms and enterprise AI tooling have become the default starting point. In housing, this often shows up through tools like Claude, OpenAI or Microsoft’s AI stack. These systems are powerful, flexible and improving rapidly. They solve for capability, but they do not solve for coordination. Many larger enterprises have been forced into building their own inhouse technology teams. Often initiated as a data aggregation team, the closed nature of the industry solutions have forced much heavier internal tech spend. For every dollar spent per unit on a closed proptech solution, there is an additional two dollars spent on derisking from lock-in and lack of data. This cost will only increase when wanting to expand the role of the internal teams to in fact solve for interoperability. Every closed vendor in an AI native expectation costs the business exponentially more in internal tooling, talent and compute. This is the second trap. The first was the proliferation of point solutions. The second is the assumption that general purpose AI can unify them. It cannot. What general purpose infrastructure provides is a foundational capability but it needs infrastructure to productionise it for enterprise level usage and governance. What is missing is a layer designed for how housing actually operates. An AI infrastructure for the housing industry. This is not a new problem. Other industries have already solved it. --- ## 05. What Other Industries Already Figured Out The structural challenge facing housing operators is not novel. The same pattern of fragmented legacy systems, proliferating point solutions, and the emergence of AI that requires cross-system coordination has already played out in legal, healthcare, and financial services. In each case, the transformative value came not from any individual AI application but from an orchestration layer positioned above the existing stack. In practice, this orchestration layer is the AI platform. ### The Legal Sector: Harvey The leading AI platform for the legal industry, Harvey, did not displace the software that law firms already used. It sat above it, routing tasks across document management systems, legal research databases, and productivity tools simultaneously. More than 100,000 lawyers across 1,300 organisations now use the platform. ### Healthcare: Abridge AI The sector’s equivalent of the property management system, the Electronic Health Record, is deeply embedded, largely closed, and controlled by a small number of dominant vendors. Abridge AI did not attempt to replace these systems. It integrates into clinical workflows, converting patient–physician conversations into structured documentation within the EHR. It is now deployed across more than 200 health systems within HIPAA constraints. ### Defense and Intelligence: Palantir Palantir’s AIP platform did not replace existing enterprise systems. It sits above them, creating a semantic model that connects data, logic, and actions across the organisation. Its U.S. commercial revenue grew 71% year-over-year in Q1 2025. The pattern is consistent across all three. The companies that captured disproportionate value did not compete with the existing software stack or even internal teams. They orchestrated it and supported industry ambitions. --- ## 06. Building the Orchestration Layer The need for an orchestration layer is no longer theoretical. The question is how it is implemented. The build versus buy conversation has always been on the table for large enterprises. The scaling cost by unit count for license fees on static software often built the case for an internal build. The case for ongoing maintenance and upgrades required dedicated teams. However, with AI orchestration, the data is always moving and agents are constantly computing. The optimisation of the platform is essential for budgets to remain predictable. So, while in-house teams can replace single purpose use cases and applications with greater ease than before, there is a challenge of staffing and underwriting a full AI platform. This needs sophisticated understanding of model selections, self hosting and row level security which are technical concepts that are often underestimated in the industry. In the absence of an industry platform, there is no choice but to take on this responsibility. However, if a platform exists for orchestration of common uses and intelligence, a partnership model can bring the benefit of “build and buy”. A platform foundation handles connectivity, orchestration and infrastructure, while internal teams build the business logic that differentiates the operation. Adopting a platform provides speed and proven infrastructure, but requires careful evaluation to avoid replicating the same lock-in dynamics the industry is trying to move beyond. The distinction is not simply build versus buy. It is where the complexity sits and who owns it over time. --- ## 07. The Action Plan for C-Suite Most operators are not starting from a blank slate. They have a PMS they cannot easily replace, a leasing stack that is three or four integrations deep, and AI pilots running in isolated workflows that have not connected to each other. The question is not what the ideal architecture looks like. It is how to move toward it from where you are. ### Action 01 — Move procurement from departmental to enterprise The fragmented stacks most operators are running today are not the result of bad technology decisions. They are the result of good departmental ones. Leasing bought for leasing. Maintenance bought for maintenance. Each tool solved the problem in front of it without accountability for what it could not connect to. AI changes the accountability structure. When workflows are expected to operate across systems, a tool that cannot participate in that model is not a departmental problem. It is an enterprise one. Procurement decisions that were previously made at the functional level now need to carry an architectural sign-off. The question is not only whether a solution solves the immediate problem, but whether it can operate as part of a stack that needs to function as a whole. ### Action 02 — Audit dependency before adding capability Before the next technology purchase, map where your data actually lives and who controls access to it. Identify which vendor relationships are load-bearing, where a commercial shift or an API change would break an operational workflow. That audit will tell you more about your AI readiness than any RFP process. Once that picture is clear, operators face a decision about how to build toward the orchestration layer. That decision should follow the audit, not precede it. > “AI hasn’t taken hold the way many expected. What the industry has built so far is a shelf of one-offs, valuable in isolation, but difficult to scale or expand.” > > ### Action 03 — Build the foundational layer on open infrastructure The foundational layer of the stack should sit on platforms whose openness is structural rather than strategic. Microsoft Azure and Fabric, MS365, Twilio and their equivalents are the enterprise infrastructure that allow for interoperability. They have established ecosystems, documented APIs and no incentive to restrict how data is used. The data warehouse, communication infrastructure and identity layer belong here. Industry-specific platforms then connect to this foundation as a purpose-built layer, extending what enterprise infrastructure cannot do on its own. ### Action 04 — Choose an industry AI platform or build one The question is no longer whether this layer will exist. It is who will build it, and how operators should evaluate it. The most durable architecture is one where your own data infrastructure sits at the centre and every vendor connects to it as a spoke. This does not require replacing your PMS. It requires ensuring that data flows out of it on your terms, into infrastructure you own. This only works if the operator remains the hub. A vendor that cannot deliver clean data and receive instructions as part of a wider system is a dependency with an exit cost, not a long-term partner. The orchestration layer is not a tool you buy once. It is the AI infrastructure for the housing industry, compounding over time as workflows mature and operational logic accumulates. Operators face a clear choice: build this capability internally or work with a platform designed to provide it. What is not viable is continuing to add point solutions and assuming integration partnerships will hold. The operators who move first on architecture will not just run more efficient operations. They will be the ones with the data infrastructure in place when AI capabilities make the gap between them and everyone else difficult to close. The defining criterion is interoperability: whether the platform can operate across systems, not within one. --- The housing industry is at a similar point in its evolution to legal, healthcare and financial services when those sectors established the orchestration layers that now define them. In each case, the operators who recognised the shift early and made decisions accordingly were the ones who captured the value. These industries chose to partner with an AI Platform partner but audit them for interoperability. The same choice is now in front of housing.

Interested in AI platforms for the industry? Contact us

## The Platform Illusion: Separating Architecture from Hype URL: https://www.travtus.com/whitepapers/the-platform-illusion Across the multifamily industry, every vendor now claims to offer a “platform.” Yet few of these systems can sustain even a modest AI workflow. The term has become a convenient label for products, bundles of tools, or integrated suites that promise simplicity while often adding complexity. This overuse has created an illusion of modernisation. Beneath the surface, many stacks remain fragmented. A true platform is not a set of features. It is an operating layer that connects people, data and workflows in a way that allows learning and adaptation. In an AI-driven world, that distinction matters. Artificial intelligence cannot function effectively without systems designed for interoperability, data continuity and openness. The companies that win in the next decade will not be those with the cleverest algorithms but those with the strongest foundations.

Download the PDF

## What a Platform Really Is — and Isn’t > “A platform is an environment you can build on, where workflows, data and partners integrate seamlessly. It’s not just what the vendor provides; it’s what the ecosystem can create.” > > In technology, a platform is an extensible foundation that others can build on. It enables integration and innovation across tools and teams. A product automates; a platform amplifies. True platforms are multilingual. They can connect to different systems and “speak” to diverse technologies without forcing conformity. The best analogy is iOS or Google’s marketing stack, where shared services, open interfaces and developer participation extend value continuously. Most platforms in multifamily technology fall short of that ideal. They resemble walled gardens. They may allow some integrations but only with preferred partners. Their structures are optimised for control rather than scale. > “We didn’t realise we were buying a product roadmap instead of a platform. Now we’re stuck waiting for quarterly releases to fix basic issues.” > > > “I don’t see too many groups using the word ‘platform’ in a way that feels intentionally misleading or disingenuous but I do see a lot of products calling themselves platforms when they’re really just well-packaged feature sets, or even a grouping of various products. > > What seems to be missing most often is interoperability. They don’t allow data or actions to move fluidly between teams or different workflows. A platform should enable cross-pollination; a ‘faux-platform’ traps you in silos with prettier branding.” > > Bundles can seem efficient at first, but they create dependency and complexity over time. A real platform delivers freedom of choice. It lets new tools connect easily, allows data to move freely and enables intelligence to build across systems. ## The System Problem: People as Middleware Across organisations, technology gaps are often filled by people rather than software. Teams reconcile data manually, copy information into spreadsheets and rely on meetings to translate insights between systems. As one industry leader put it: > “Right now, our workflows depend on people acting as middleware between disconnected tools.” When humans become the connectors, technology loses its purpose. Efficiency turns into administrative burden. The next wave of systems must automate that connectivity. A true platform allows information from one workflow — leasing, accounting or maintenance — to inform another automatically. This is the essence of system thinking: value emerges not within products but between them. AI exposes the cost of poor integration. Machine learning models depend on continuous, structured and reliable data. When those connections break, intelligence collapses. Interoperability is therefore not a luxury; it is the condition for progress. ## Data Portability: The Core of Intelligence Every conversation with executives led to the same conclusion: **data portability defines a platform**. Without the ability to move, combine and reuse data across systems, even the most advanced automation remains shallow. When data cannot move, organisations fill the gap with manual work. They maintain shadow systems and duplicate logic across tools. These hidden costs erode both efficiency and morale. Platforms remove that friction. They create continuity so information travels cleanly and instantly between systems. Each new data source strengthens the organisation’s collective intelligence. For AI, portability is oxygen. Models learn by detecting patterns across multiple contexts. That is impossible without unified, accessible data. A closed system might automate a process, but it will never create organisational learning. The measure of a platform is not how many features it has but how easily data flows through it. > “Portability is what separates a product that automates from a platform that empowers.” > > ## Architecture Over Roadmaps Technology evolves in stages. First comes the **product**, built to solve a specific problem. Next, the **suite**, which combines related tools under one vendor. Finally, the **platform**, an open foundation that links data, workflows and users across the business. Most systems never reach that final stage because they lack the architectural openness to do so. Without shared data models and transparent interfaces, suites harden into silos. Forward-thinking operators design for openness from the beginning. They avoid lock-in, adopt integration standards and retain ownership of their data. Some push information into analytics layers; others build orchestration tools that sit above existing products. Both approaches recognise that architecture matters more than any vendor roadmap. Governance makes this discipline sustainable. Many firms have created technology steering committees or centres of excellence to maintain integration standards, protect data quality and ensure software choices align with business strategy. For third-party managers, the barriers are often structural. Data ownership sits with property owners, fragmenting visibility and limiting scale. Yet progress is possible. One executive, speaking privately, said: “You can’t always centralise data, but you can centralise principles.” Clear standards for privacy, consent and data exchange allow even federated organisations to behave like platforms. For AI, such governance is more than good housekeeping. Algorithms trained on inconsistent or incomplete data produce unreliable results. Disciplined architecture creates trust in the intelligence built on top of it. ## The Economics of Platform Thinking Elegant platforms hide heavy costs. Real interoperability requires investment in data ingestion, tagging and identity resolution, along with the people who maintain them. As one operator explained anonymously: > “Every data point has a cost. The more you integrate, the more you pay.” This is the paradox of openness. Flexibility demands more effort before it yields returns. Some organisations choose to simplify, feeding high-quality data into a few core systems rather than building a full ecosystem. Others invest early, believing that automation and analytics will repay the expense. Either path demands patience. The payoff from platforms arrives in years, not quarters. Point tools produce quick wins; open systems deliver long-term advantage. In capital-disciplined environments, the question has shifted from **Can we afford a platform?** to **Can we afford not to have one?** The hidden costs of fragmentation — manual reconciliation, lost data, inconsistent reporting — accumulate faster than any software subscription. ## AI as the Forcing Function Artificial intelligence is redefining what a platform must be. Traditional tools execute instructions; AI learns from interaction. That difference makes architecture a strategic dependency. AI systems need structured, connected and real-time data, as well as context across workflows and feedback loops that refine their performance. Without those elements, even the best model becomes a static feature. AI is, in effect, the stress test for platform truth. Closed products can mimic intelligence for a while, but they cannot evolve. True platforms support continuous learning. They provide the environment where algorithms improve as they encounter more data and human feedback. The next generation of enterprise technology will be judged by this operability. Intelligent orchestration will replace narrow automation. AI will coordinate leasing, marketing, maintenance and finance in real time — but only if those systems share a common foundation. For companies building AI, this is decisive. The quality of a model depends less on clever code and more on the data infrastructure behind it. AI becomes not a feature but the outcome of sound architecture. Operators should reverse the usual question. Instead of asking what AI can do for a platform, they should ask whether the platform can sustain AI. The answer will reveal whether they have built a foundation or a facade. ## Platform Maturity: Signs of the Real Thing How can operators tell a true platform from an imitation? The clues lie in architecture rather than appearance. | Signal | What it looks like | | --- | --- | | **Openness** | Public, well-documented interfaces that anyone can connect to. | | **Data Portability** | Clear ownership terms and real-time export options. | | **Interoperability** | Compatibility across vendors, including legacy systems. | | **Governance** | A mechanism to enforce integration standards and data hygiene. | | **Extensibility** | A marketplace or module framework that invites innovation from others. | These are not only technical qualities but economic ones. Open systems reduce switching costs and encourage competition around them. Closed systems accumulate dependence and debt. The former generate resilience; the latter consolidate risk. > “Ideally, every feature would be a painkiller, not a vitamin — solving real bottlenecks rather than adding surface-level ‘nice to haves.’ It would be modular enough to grow with businesses, open enough to integrate easily, and disciplined enough to avoid bloat.” > > ## The Human Dimension Platform transformation is as much cultural as technical. Organisations that treat platforms as strategic infrastructure — rather than as vendor contracts — achieve faster returns and stronger adoption. That shift demands executive alignment. Platform thinking crosses departmental lines. It requires cooperation between IT, operations and finance. Without leadership sponsorship, even the most elegant system fragments under competing priorities. Many firms now operate “platform councils” or “centres of excellence” to manage both the politics and the architecture. Their job is not simply to approve purchases but to define the rules of the ecosystem: data openness, integration discipline and API standards. In multifamily, where ownership structures are fragmented and governance uneven, this cultural readiness may decide who thrives in the AI era. Success will favour those who build coherence, not those who buy more tools. ## Conclusion: Architecture as Destiny The industry’s fascination with “platforms” hides a more important question: which of these systems can truly host intelligence? AI exposes weakness instantly. It fails when data is trapped, when integrations are shallow and when workflows lack continuity. Marketing may sell “AI-powered” products, but sustained intelligence depends on clean data, open architecture and disciplined governance. The future belongs to organisations that treat platforms as living infrastructure rather than marketing claims. Multifamily does not need more vendors calling themselves platforms. It needs systems that act like them: interoperable, transparent and ready for AI. A platform is not something a company buys; it is something it builds — deliberately, openly and with intelligence in mind. Those who understand that will not just adopt AI. They will operate intelligently. The illusion ends when the industry stops buying promises and starts demanding architecture. --- ## Expert Insights from Newmark RF ### Q1: When you hear the term “platform,” what does it mean to you? A platform is not just a system, it is a technology ecosystem that underpins the entire digital strategy of an operator. It connects, integrates, and scales technology across the organization. It enables seamless collaboration, unified data, and the extensibility to evolve with the business. A true platform provides: - A foundational ecosystem for multiple applications and services - Shared capabilities that reduce redundancy - Open integration frameworks that connect internal teams and third-party tools - The ability to innovate at scale without constant reinvention ### Q2: Is the industry working with systems that claim to be platforms but feel more like pipelines or funnels? What’s missing? Yes, it is a common challenge. Many systems are marketed as “platforms”, but operate more like pipelines, closed, linear tools designed around one way of working. What is usually missing: - **True openness:** Limited APIs, lack of third-party integration, and restricted customization - **Configurability:** Inability to adapt workflows to different operating models or business processes - **Scalability:** Struggles to grow or evolve with the portfolio - **Data interoperability:** Siloed information and fragmented user experiences ### Q3: What role should a platform play in the tech stack of a modern operator? A platform should serve as the strategic backbone of the technology stack orchestrating how data, processes and tools come together across the organization. The platform should allow operators to move faster, work smarter and scale seamlessly. It should: - Centralize data and insights across property, asset, development and financial operations - Support flexible integration with both legacy systems and emerging Proptech solutions - Enable automation and decision intelligence across the organization - Power innovation without disrupting core operations ### Q4: In your view, what are the key capabilities or qualities that make a system a platform, not just a product? A product solves a specific function. A platform creates a foundation for many solutions to work together, adapt and grow as needed by the business. Key platform qualities: - **Interoperability:** Integrates seamlessly with third-party and legacy systems - **Extensibility:** Allows new modules, features, or tools to be added over time - **Configurability:** Adapts to different workflows and business models - **Data unification:** Centralizes and normalizes data across functions - **Scalability:** Supports operational growth without rework or disruption - **Security and compliance:** Built-in governance and data protection ### Q5: If you could design your ideal platform from scratch, what would it include and what would it avoid? | Must-haves | What to avoid | | --- | --- | | • Open APIs and an integration-first architecture
• Modular design, so components can be used independently or together
• User-centric experiences with self-service and configurability
• Real-time data access and embedded analytics
• Robust security, compliance, and governance controls
• Support for custom extensions or partner apps | • Closed ecosystems that block third-party integration
• Rigid workflows that require heavy customization
• Siloed data models that limit enterprise visibility
• Vendor lock-in strategies that reduce flexibility | ### About Newmark RF [**Newmark RF**](https://www.realfoundations.net/) is a global professional services firm focused on helping real estate companies that develop, own, operate, or invest in real estate make smarter, more profitable decisions. Our client portfolio represents a combined asset value of over $10 trillion, covering more than 10 billion square feet and over 7 million distinct residential units across the world. Newmark RF is a trusted advisor helping the industry navigate the evolving platform landscape. With deep operational and technical expertise, Newmark RF helps: - Assess current-state ecosystems, identifying fragmentation and duplication - Define future-state architecture, aligned with strategic goals and growth plans - Guide platform selection, using structured evaluations across technical, operational, and financial criteria - Ensure implementation success, through change management, integration support, and training Newmark RF ensures that technology investments are practical, scalable, and aligned with business outcomes, turning platforms into real performance enablers. ## The Stack Reset: From Chaos to Clarity in Multifamily Real Estate URL: https://www.travtus.com/whitepapers/the-stack-reset Now, the industry is entering a new era — one defined by restraint, replatforming, and a clear-eyed focus on what actually works. **This is the story of The Stack Reset: the deliberate move away from noise and novelty, and toward tech ecosystems that are leaner, smarter, and built to last.**

Download the PDF

In 2020, everything in property management went digital overnight. Leasing agents, once the face of the community, went remote. Maintenance teams adapted to social distancing. Owners demanded new metrics, new tools, and faster decisions. The industry met the moment — but it did so in a rush. Technology flooded in: virtual tour software, self-guided leasing, mobile maintenance apps, prospect communication tools. Much of it worked. Some of it didn’t. But very little of it was part of a unified strategy. > “We went from a tech desert to a tech downpour — and didn’t have a raincoat.” > > Now, five years later, the landscape is cluttered. Many operators find themselves with dozens — or even hundreds — of tools layered into their portfolios. No one quite knows how they got there. And everyone’s wondering how to make sense of it all. Much of the confusion stems from how these tools entered the organization. Marketing teams often purchased leasing or communication tools independently, eager to improve prospect engagement. Operations teams experimented with automation to handle mounting tasks. Owners frequently introduced vendors with ties to their other assets or investment portfolios. And IT departments — stretched thin — had little oversight or enforcement power. The result wasn’t just a stack — it was a patchwork. Teams now find themselves navigating tools they didn’t select, supporting platforms they didn’t scope, and stitching together workflows they didn’t design. As one executive put it, “No one owns the whole picture. Everyone owns a piece of the pain.” ## A Stack That Grew in the Shadows Tech stack sprawl didn’t happen because operators were careless. It happened because everyone was trying to solve urgent problems in real time — and departments moved fast. Marketing brought in leasing tools. Owners pushed their preferred vendors. On-site teams requested scheduling apps. Meanwhile, IT — already stretched — often had little say. One operator revealed they had more than 140 vendors per property, many of them doing similar jobs in different regions. > “Most organizations don’t even know what they’re using anymore. There’s no inventory, no visibility, and no ownership.” > > And integration, the promise that held it all together, often failed to deliver. While APIs were touted in sales decks, actual data flow was messy. Some integrations pulled in incomplete data. Others failed to sync at all. One executive described their experience bluntly: “Our teams are spending their time checking whether the integrations worked — not actually using the systems.” **Who bought what?** | Team | What they bought | | --- | --- | | **Marketing** | CRMs, leasing automation, prospect engagement | | **Operations** | Task workflows, communication tools, maintenance platforms | | **Ownership** | Asset-specific tools, BI systems | | **IT** | Playing catch-up with oversight and integration | **The result:** A tech stack without a clear owner — and a lot of disconnected effort. ## Three Paths to Clarity The frustration has led to a broader movement — a deliberate reevaluation of the tech stack. Across the industry, operators are pursuing what many are calling The Stack Reset. But while the end goal is the same — fewer, better-integrated, more usable systems — the approaches differ. ### 1. Vendor Consolidation For some, the reset means cutting aggressively. At RPM, Scott Pechersky, Chief Technology Officer, is reducing tools not just for cost, but for cohesion. This approach favors fast wins — shrinking contracts, eliminating redundancy, and lightening the load on teams. But it comes with tradeoffs: in some cases, specialization is lost. Some tools “almost work” for everyone, but not quite anyone perfectly. > “We’re consolidating from more than 40 platforms to about 15. You get better training, better adoption, and a better handle on what’s working.” > > ### 2. Platform Re-Centering Others are doubling down on their Property Management System (PMS) — not because it does everything well, but because it’s the one tool everyone touches. > “Core platforms like Yardi do a lot — property management, GL, job costing. But when we add something new, it has to integrate cleanly. That’s non-negotiable.” > > For these teams, the goal is not perfection — it’s predictability. If a solution can live within or directly adjacent to the PMS, it gets a hearing. If not, it’s often a pass. Still, it’s worth noting: most existing PMS platforms suffer from deep architectural rigidity, caused by: - Scope creep across task management, communication, and operational modules - Bundling of unrelated functionality into core schemas - Static data models stretched to serve diverse use cases This rigidity explains why some operators pursue a third path. ### 3. Rebuilding Internally A third approach — used more selectively — is to build from within. These teams are not layering on tools; they’re building infrastructure. Internal platforms, APIs, and data layers that support both day-to-day operations and long-term flexibility. This approach requires strong internal capabilities. But it’s appealing to operators who want control over iteration, data accuracy, and future extensibility. “When you build something yourself,” one executive explained, “you’re not just a user. You’re an architect. That changes how the stack evolves.” ### The Venture Capital Hangover Some of the challenges now surfacing were seeded by the funding environment of the early 2020s. Venture-backed proptech startups were often advised to “find a wedge” — solve a narrow pain point, raise money, and then expand. The problem? Many never expanded. Their MVPs stayed minimal. The integrations never matured. The platforms never arrived. Operators bought in — hoping the roadmap would deliver — but found themselves stuck with thin tools that couldn’t scale. “A lot of easy money came with bad advice,” one executive said. “There were too many solutions solving one problem — and none solving the whole picture.” As funding dries up, some of those vendors are disappearing. Others are freezing development. And operators are left holding the bag — along with the maintenance contracts. ## AI: A Capability Waiting for a Platform Over the past two years, AI has emerged as the most talked-about layer in multifamily tech. And in most cases, the early results have been encouraging. Operators have piloted AI to automate document verification, predict renewals, score leads, and manage internal workflows. The tools work. The promise is real. But something is off. AI hasn’t taken hold the way many expected. What started as energy has turned into inertia. Many operators are left with disconnected pilots — valuable in isolation, but difficult to scale or expand. “Everyone’s excited about what AI can do,” said one senior ops lead. “But what we’ve built so far is a shelf of one-offs.” At the core of the issue is architecture. AI, unlike earlier software cycles, doesn’t succeed just by being clever — it succeeds when it’s embedded, flexible, and broadly connected. It needs access to data, clear event flows, and room to iterate. But most data is hidden behind walls of vendor systems and not accessible to operators. The challenge is that few operators have invested in a platform approach towards AI. And vendors that claim to offer one often struggle with the same limitations as everyone else: brittle integrations, poor data exchange, and inflexible schemas. So instead of transformation, most teams are getting repetition — trying to make smart tools work in systems that weren’t designed to be dynamic. Still, the desire hasn’t gone away. In fact, it’s growing. Across the interviews, operators described AI less as a novelty and more as an expectation. Not just automation — but adaptation. Not just savings but scale. The question now is: who will build the infrastructure to support it? And how long can the industry afford to wait? Because while the first wave of AI in multifamily has already happened, the second one — where it actually changes how work gets done — is still waiting for a place to land and the right sponsors. ### What Would a Real AI Platform Look Like? Most of the AI tools in multifamily today resemble apps — not platforms. They automate single tasks, run one model, or replace one decision — but they don’t scale across functions. They don’t learn from outcomes. And they rarely persist beyond the pilot. A real AI platform is something else entirely. **An AI platform in multifamily would need to:** - Sit across operational, financial, and resident data - Allow users to configure their own workflows, not just consume fixed outputs - Enable teams to test, refine, and scale AI use cases — without ripping out core systems - Treat AI not as a product, but as infrastructure This is not something that can be patched together through MVPs. It requires deliberate investment, deep architectural vision, and a data strategy that puts operators — not just vendors — in control. ![Where Travtus sits in the multifamily competitive landscape](/assets/whitepapers/the-stack-reset__Competitive_Landscape_Diagram_For_Web.png) ## Rebuilding with Intention This time, operators are asking harder questions: - **What are we using?** - **Who owns it?** - **How do I get access to all the data?** There’s no single answer. But there is a new consensus: the stack must serve the work — not the other way around. The Stack Reset isn’t about saying no to innovation. It’s about saying yes to stability, visibility, and scale. It’s about remembering that technology, at its best, disappears into the background — and lets teams focus on the residents, not the software. After years of chaos, the industry isn’t chasing more tech. It’s chasing clarity. > “In three to five years, AI will help solve the integration problem. And when that happens, best-of-breed might come back. But until then, we’re betting on systems we know we can run.” > > --- # Articles ## Blueprint 2026 Sessions: The Seven That Show Where Multifamily AI Buying Is Headed URL: https://www.travtus.com/resources/blueprint-2026-sessions-multifamily-ai Date: 2026-09-15 Summary: A curated guide to seven Blueprint 2026 sessions on multifamily AI, with day, time, room, speakers and the one question worth asking in each. The agenda has stopped asking what AI does and started asking what it costs, what it replaces and who builds it. > **The seven Blueprint 2026 sessions worth your time run September 22 to 24 in Las Vegas, and they share one theme: AI is now a cost, consolidation and ownership question, not a capability question. Start with "What We Deleted This Year" (Tuesday, 12:45pm, Palazzo C), which is about retiring tools rather than adding them.** A conference agenda is a leading indicator. Not of what an industry is doing, but of what it has finally admitted it is worried about. Read Blueprint 2026 that way and the shift is hard to miss. **Of the seven sessions below, not one names an AI capability.** Four name money: value, dollars, ROI, NOI. One names subtraction. One names owning the whole chain. One names teaching a system your own property data. Zero name a feature. That distribution is the story. ## The seven sessions, at a glance Bookmark this table. It is the whole guide on one screen, and the last column is the part worth having in your pocket when you sit down. | When and where | Session | Speakers | The one question to ask in the room | | --- | --- | --- | --- | | **Tue 22**
12:45 to 1:20
Palazzo C, Track 2 | **What We Deleted This Year** | Chelsea Kneeland, *Multifamily Insiders*
Carol Enoch, *Bridge Partners*
Dalia Kalgreen, *Unified Residential*
Steve Boyack, *JVM Realty*
Devang Patel, *Morgan Properties* | What did you delete that *you* had championed yourself? | | **Tue 22**
1:25 to 2:00
Palazzo D, Track 1 | **Underwritten: How to Value Technology in SFR** | Reshma Block, *Compass Consulting*
Joe Polverari, *Pure HomeRiver*
Mahesh Shetty, *ILE Homes*
CJ Halabi, *Pretium* | How do you underwrite a platform that has not shipped its roadmap yet? | | **Tue 22**
3:15 to 3:50
Palazzo C, Track 2 | **From Data to Dollars** | Melissa Fagan, *RET Ventures*
Bobbi Steward, *Revyse*
Danny Bin, *Principal Asset Management*
Scott Pechersky, *ECL Prop Solutions* | What share of the insights your systems generate is ever acted on? | | **Tue 22**
3:50 to 4:25
Palazzo C, Track 2 | **AI in Action: Turning Operational Data into Real ROI** | Mariana Estrada, *RPM Living*
Jameson Hartman, *RET Ventures*
Paul Seifert, *Continental*
Sorin Michnea, *SuiteSpot* | What ROI did you expect and not get? | | **Wed 23**
9:30 to 9:50
Marcello, Main Stage | **The Vertically Integrated Bet** | Christopher Yip, *RET Ventures*
Steven DeFrancis, *Cortland* | What would you *not* integrate? | | **Wed 23**
2:00 to 2:10
Palazzo D, Track 1 | **Teaching AI What the Property Actually Looks Like** | Saharsh Chordia, *Cortland* | What did the model get wrong before it was taught? | | **Wed 23**
2:00 to 2:35
Palazzo C, Track 2 | **What Does "AI-First" Have to Do With Multifamily NOI?** | Dom Beveridge, *20for20*
Douglas Pearce, *Waterton*
Aaron Ross, *Birge & Held*
Jeremy Voigtmann, *Harbor Group* | What changed about how work is assigned, not about your software? | ### Three things to plan around **Tuesday afternoon is one block, not two sessions.** From Data to Dollars and AI in Action run back to back in Palazzo C. Take a seat at 3:10 and stay in it until 4:25. No room change, no gap. **Tuesday lunchtime is a five-minute turnaround.** What We Deleted ends 1:20 in Palazzo C, Underwritten starts 1:25 in Palazzo D. Doable if you sit near an aisle and do not stop to talk on the way out. **Wednesday at 2:00 you have to choose, but you can partly cheat.** Both sessions start at once in different rooms. Teaching AI runs only ten minutes, so you can take it in full and walk to Palazzo C, at the cost of missing the opening of the NOI panel. Technical, take the ten minutes and walk. Commercial, stay put. ## Why each session earns its slot ### "What We Deleted This Year": which Blueprint 2026 session covers retiring software and consolidating platforms? **Tuesday 22 September · 12:45pm to 1:20pm · Track 2 · Palazzo C** **Speakers** - Chelsea Kneeland, Innovation Insider, Multifamily Insiders - Carol Enoch, Asset Manager, Bridge Partners - Dalia Kalgreen, VP, Unified Residential - Steve Boyack, Chief Operating Officer, JVM Realty - Devang Patel, SVP, Head of Technology, Morgan Properties This is the single most important session on the agenda, and it is on at lunchtime on day one, which tells you the industry has not yet worked out how important it is. The framing in the official description is unusually direct. For years the industry focused on adding technology. Leading owners are now asking what they can eliminate: retiring legacy software, consolidating platforms, reducing logins, cutting costs, and building internal solutions instead of buying another product. That last clause is the one to sit with. It is a build-versus-buy statement made by owners, on the record, on a main-track stage. What makes it credible is the composition of the panel. An asset manager, a chief operating officer, a head of technology and a residential leader are four different accountabilities, and they usually disagree about software. An asset manager cares what the line item does to the model. A COO cares what a rollout does to the field. A head of technology cares what the integration debt looks like in three years. When four seats that normally pull in different directions are booked into one room under the word "deleted", the disagreement has already been resolved somewhere, and this session is where it becomes public. **The question to ask in the room:** ask them what they deleted that they had championed themselves. Anyone can retire an inherited tool. The interesting number is how many of the things they switched off were things they had personally signed for, and what it cost politically to admit it. ### "Underwritten: How to Value Technology in SFR": how do you put a value on technology in single-family rental? **Tuesday 22 September · 1:25pm to 2:00pm · Track 1 · Palazzo D (SFR/BTR track)** **Speakers** - Reshma Block, Principal & Founder, Compass Consulting - Joe Polverari, Co-Founder & CEO, Pure HomeRiver Property Management - Mahesh Shetty, CEO, ILE Homes - CJ Halabi, Managing Director and the Head of Digital Product, AI & Business Innovation, Pretium Sitting on the single-family track means most multifamily attendees will skip it. That is a mistake, because single-family rental is where the technology valuation question gets asked most honestly. The reason is structural. Scattered-site portfolios cannot hide a bad technology decision behind on-site staffing the way a dense multifamily asset can. If your systems are poor across six thousand homes in nine markets, there is no property manager standing in a leasing office absorbing the difference. The cost surfaces immediately. So single-family buyers were forced into rigor about what a platform is actually worth earlier than the rest of the industry, and the word in the title is "Underwritten", not "evaluated". That is a specific claim: that technology now shows up in the model, defended, rather than in an operating expense line nobody interrogates. The mix of seats matters again. Two chief executives, a founder-principal and a managing director whose remit explicitly spans digital product, AI and business innovation. That is a valuation conversation, not a product conversation. **The question to ask in the room:** ask how they underwrite a platform that has not shipped its full roadmap yet. Every honest answer to that question is really an answer about how much they trust the vendor, and it is worth hearing said out loud. ### "From Data to Dollars": how do you turn operational insight into value you can actually bank? **Tuesday 22 September · 3:15pm to 3:50pm · Track 2 · Palazzo C** **Speakers** - Melissa Fagan, Vice President, RET Ventures - Bobbi Steward, CEO, Revyse - Danny Bin, Principal Asset Management, Head of Private Market IT - Scott Pechersky, Founder, ECL Prop Solutions The word doing the work in this title is "Immediate". Not eventual, not at scale, not in year two. The industry has spent a decade building dashboards that produce insight nobody acts on, and the honest version of that failure is that an insight which does not change what somebody does on Thursday morning is a cost, not an asset. Notice the venture seat on the panel. Capital allocators sit on the other side of this question from buyers, and they see the pattern across a portfolio of companies rather than one estate. When investors and buyers are put on the same stage to argue about the gap between insight and value, you get a better conversation than either group has on its own, because the investor knows which claims did not survive contact with a real portfolio. **The question to ask in the room:** ask what percentage of the insights their systems generate are ever acted on. Almost nobody measures this. The people who do measure it are the people worth listening to. ### "AI in Action": how do you get real ROI from operational data? **Tuesday 22 September · 3:50pm to 4:25pm · Track 2 · Palazzo C** **Speakers** - Mariana Estrada, Chief Strategy Officer, RPM Living - Jameson Hartman, Industry Principal, RET Ventures - Paul Seifert, EVP Operations, Chief Legal Officer, Continental - Sorin Michnea, CTO, SuiteSpot This runs immediately after the previous session, in the same room, and the pairing is deliberate enough that I would treat the two as one ninety-minute block rather than two sessions. The seat worth watching is the combined operations and legal role. Those two accountabilities are rarely held by one person, and when they are, the AI conversation gets sharper, because the same individual owns both the upside of moving faster and the exposure of getting it wrong. Most ROI panels are staffed entirely by people who own only the upside. This one is not. The word "Real" in the title is doing defensive work. It concedes that a lot of the ROI claimed in this category over the past three years was not real, and it invites the panel to say why. That concession is progress. **The question to ask in the room:** ask them to describe the ROI they expected and did not get. The delta between the business case and the outcome is the most useful number in the room, and it is the one almost never published. ### "The Vertically Integrated Bet": is vertical integration the answer to a fragmented resident experience? **Wednesday 23 September · 9:30am to 9:50am · Main Stage · Marcello Ballroom** **Speakers** - Christopher Yip, Partner & Managing Director, RET Ventures - Steven DeFrancis, Founder & CEO, Cortland Twenty minutes, main stage, two people. That format is a signal on its own. Main-stage slots that short are booked when the organizers believe a single argument is worth hearing without a panel diluting it. The word in the title is "Bet", which is more honest than most main-stage framing. Vertical integration is a capital-structure decision before it is a resident-experience decision, and "wall to wall" is a claim about owning the whole chain rather than assembling it from vendors. Whether or not you think that bet is right for your own balance sheet, it is the same question as the deletion session asked from the opposite direction. One asks what you can stop buying. This one asks what you should own outright. If you attend only one session on this list, and you sit on the investment side rather than the running-the-portfolio side, make it this one. **The question to ask in the room:** if there is audience question time in twenty minutes, ask what they would not integrate. The boundary of the bet is more instructive than the bet. ### "Teaching AI What the Property Actually Looks Like": how do you teach AI your own property data? **Wednesday 23 September · 2:00pm to 2:10pm · Track 1 · Palazzo D** **Speaker** - Saharsh Chordia, Vice President, Strategy and Analytics, Cortland Ten minutes, one speaker. It is the shortest thing on this list and, for anyone technical, possibly the most interesting. Almost every AI conversation in this industry assumes the model arrives already knowing what a property is. It does not. A model that has never seen your unit mix, your amenity naming, your maintenance categories or the way your teams actually describe a problem in free text will produce answers that are fluent and wrong. Closing that gap is not a procurement exercise. It is the work of teaching a system your own context, and whoever does that work ends up owning the resulting capability. That is the "who builds it" leg of the agenda's argument, and it is the leg most buyers have not thought about yet. If a vendor teaches its model using your data and the capability stays theirs, you have funded someone else's product. If you teach it and the capability stays yours, you have built an asset. The session title is about property data. The subtext is ownership. **The question to ask in the room:** ask what the model got wrong before it was taught, and what the teaching actually consisted of. Specifics here separate real work from a slide. ### "What Does 'AI-First' Have to Do With Multifamily NOI?": does an AI-first operating model move NOI? **Wednesday 23 September · 2:00pm to 2:35pm · Track 2 · Palazzo C** **Speakers** - Dom Beveridge, Founder, 20for20 - Douglas Pearce, EVP, IT, Waterton - Aaron Ross, President, Birge & Held - Jeremy Voigtmann, Managing Director, Harbor Group The title is phrased as a challenge rather than a claim, and the quotation marks around "AI-First" are not accidental. They put the term itself on trial, which is the right instinct. "AI-first" has been used loosely enough over the past two years that it now needs defending before it can be used. The panel is built to test it. A president, a managing director, an EVP of IT and an independent industry analyst is a combination that will not let a vague answer stand. And the connection the title proposes is the hard one: not whether AI is useful, but whether an operating model organized around it produces a different NOI than one that is not. Those are genuinely different questions. Plenty of companies have useful AI and an unchanged operating model, which is precisely why their NOI looks the same as it did before. **The question to ask in the room:** ask what they changed about how work is assigned, not what they changed about their software. An operating model that has not changed who does what has not changed. ## What this agenda tells you about how AI gets bought next year The 2024 pattern was additive. Find a discrete problem, buy a tool aimed at it, integrate it, repeat. It produced real wins, and it produced the condition every one of these sessions is now responding to: a stack of narrow tools, each defensible on its own, that collectively cost more than anyone modeled and cannot answer a question spanning two of them. The 2026 pattern, read off these titles, is subtractive and ownership-led. Establish what a technology is worth on the model before you buy it. Ask what it replaces, not only what it adds. Judge an insight by whether it changes what someone does. Decide deliberately what you own and what you rent. And treat teaching a system your own context as capability you are building, not a service you are consuming. That is a portfolio-management argument that happens to have technology as its subject, which is why these panels are staffed with asset managers, presidents and chief operating officers rather than product leads. If you want the longer version of two of these threads, we have written them up separately: the structural case for [an AI platform versus point solutions in multifamily](/resources/platform-vs-point-solutions), and a practical framework for [how to evaluate an enterprise AI platform for real estate](/resources/how-to-evaluate-enterprise-ai-platform-real-estate) that covers the underwriting and build-versus-buy questions sessions two and one put on stage. The useful thing about going to Blueprint with this frame is that it changes what you do on the exhibition floor. If your list of questions is "what does it do", you will have thirty pleasant conversations and remember none of them. If your list is "what does this replace, what does it cost fully loaded, and who owns the capability once my data has taught it", you will have four difficult conversations and come home with something you can take to your investment committee. The full programme is on the [Blueprint agenda](https://blueprintvegas.com/agenda/). If the deletion question is the one you leave Vegas with, start with the two pieces linked above and bring the answer to your next investment committee rather than your next demo. Travtus is the everyday AI platform for multifamily. Build your own AI, for your properties, in a sentence. **Yours to Build In a Sentence.** ## AI in Real Estate: Why the Operating Model Has Become the Constraint URL: https://www.travtus.com/resources/ai-real-estate-operating-model-the-district-2026 Date: 2026-09-03 Summary: As AI moves into daily work across the UK, US and European markets, the binding constraint has shifted from technology to organizational design. Analysis ahead of The District 2026 in Madrid. > **The constraint on AI in real estate is no longer the technology. It is the operating model. AI collapses the time it takes to reach a decision, but approval thresholds, reporting lines and job descriptions still assume the old speed, so organizations acquire a capability they have no mechanism to use. The organizations that benefit re-cut decision rights first, assign accountability explicitly, and build on data they own.** ### In brief - The AI conversation in real estate has converged internationally. Across the UK, the US and continental Europe, the question has moved from whether the technology works to how decision rights, roles and accountability should be redesigned once it does. - Most stalled AI programs in real estate are not technology failures. They are organizational design failures: decision latency collapses while approval paths, reporting lines and job descriptions continue to assume the old speed. - The strategic question for owners and investors is one of ownership. Licensed software encodes a vendor's view of how a business should run. Intelligence built on an organization's own operating data is a proprietary asset that survives a change of system. ## A conversation that has converged For most of the last decade, AI in real estate was discussed differently in each major market. In the United States the framing was productivity. In the United Kingdom it was regulatory and compliance-led. Across continental Europe it was predominantly a data and governance question. Those framings have converged with unusual speed. The question now being asked in Madrid, London and Dallas is substantially the same: as AI becomes embedded in daily work, who decides, how fast, and who is accountable for the outcome. The technology debate has largely settled. The organizational debate has not, and it is where the material disagreement in the sector now sits. That is the subject of **AI-Native Operations and Teams**, a session in the Artificial Intelligence track at The District 2026 in Madrid on 22 September, featuring Anjli Carlucci, AI Engagement and Adoption Lead at Grosvenor Group, and Tripty Arya, Founder and CEO of Travtus. The event's own abstract frames it precisely: as AI becomes embedded in daily operations, organizations are being forced to rethink how decisions are made, and roles, workflows and accountability are evolving in practice. The inclusion of accountability is notable. At a capital markets event, an AI session description now places accountability alongside workflow rather than alongside efficiency or cost. That is a meaningful shift in how the industry is framing the problem. ## Why AI programs stall AI adoption in real estate is still commonly treated as a procurement decision: which tool, which vendor, which integration. In Travtus's experience working with portfolios covering more than 500,000 homes, that framing is the most reliable predictor of a program that underdelivers. The reason is structural. An organization is, in effect, a set of agreements about who is permitted to decide what, on what evidence, within what time. AI renegotiates those agreements implicitly. It changes the speed and confidence with which a decision can be reached, without changing the authority structure built around the old speed. The result is a capability the organization cannot use. ## Three shifts reshaping how real estate organizations work ### 1. Decision latency collapses before the organization adapts Where an analysis previously took a regional manager two days, it may now take minutes. The approval threshold, the escalation path and the role definition, however, were all designed around the two-day version. The organization gains speed it has no mechanism to spend. The practical consequence is that value accrues only where authority is re-cut to match the new latency. Organizations that deploy the capability but leave decision rights untouched typically report improved reporting and unchanged outcomes. ### 2. Roles recompose rather than disappear The work absorbed first is rarely a complete role. It is the connective tissue between roles: chasing status, re-keying data between systems, reconciling versions, establishing what was agreed and by whom. What remains is more judgment-dense. This has a workforce planning implication that is frequently missed. The residual role is materially different from the one that was hired for, and it demands a different profile and different support. Planning for a change in the composition of work is more useful than planning for a change in headcount. On the headcount question itself, restraint is warranted. The available evidence does not yet support a confident directional claim for the sector, and outcomes vary substantially by organization and operating context. Forecasts presented with certainty should be treated with caution. ### 3. Accountability has to be designed, not assumed Human oversight is widely stated as a control and less widely engineered as one. Where a system recommends and a person approves without the time, information or standing to dissent, the organization has not retained a human in the loop. It has produced a signature. Designing genuine accountability means specifying, in advance, who owns which class of decision, what evidence they are shown, what dissent costs them, and how the record is retained. This is governance work rather than technology work, and it is the most common gap in otherwise capable deployments. ## The ownership question For owners, asset managers and investors, there is a further dimension that sits above operational performance. Licensed software is another organization's intellectual property. It is configured within the boundaries the vendor permits, and the working practices it encodes leave with it at the end of the contract. What has been purchased, in substance, is a third party's opinion about how the business should run. Intelligence built on an organization's own operating data is different in kind. It encodes that organization's underwriting judgment, service standards and the specific practices in which it holds an advantage. It does not reset on renewal, and it does not transfer to a competitor buying the same product. Framed this way, the strategic question for 2026 is less which AI vendor an organization selected, and more what proprietary capability it has accumulated that an acquirer would pay for. Operating capability, historically treated as cost, becomes a component of enterprise value. ## Why residential at scale is a leading indicator Residential portfolios are a useful proving ground for this argument, and a demanding one. An asset-level view of a portfolio is a periodic abstraction, reviewed monthly or quarterly. Running homes generates a continuous stream of high-frequency decisions, the majority made by frontline staff under time pressure with incomplete information. Applying AI in that environment is harder than applying it to a periodic model, and considerably more revealing: system weaknesses surface immediately, because a resident is waiting for an answer. Lessons from that environment tend to generalize upward into asset and portfolio management. The reverse is rarely true. ## What distinguishes organizations that make the shift Across the deployments Travtus has observed, the organizations that convert AI capability into changed outcomes tend to share four characteristics: 1. **They redesign decision rights first.** The operating model changes, and the technology is fitted to it, rather than the reverse. 2. **They assign accountability explicitly.** Ownership of each decision class is named, documented and resourced, not left to be inferred. 3. **They build on their own data.** The capability reflects the organization's standards and judgment rather than a vendor's defaults, and it remains theirs. 4. **They plan for the composition of work, not a headcount target.** Role redesign and capability building are treated as part of the program rather than as a downstream consequence. ## The discussion at The District The District 2026 convenes more than 15,000 attendees from the international real estate investment market in Madrid from 22 to 24 September, having relocated from Barcelona. That the AI track now foregrounds organizational design rather than tooling is itself an indicator of where the sector's attention has moved. The open question, and the one worth pressing in the room, is one of sequence: in organizations that have genuinely changed how they work, did the operating model change first and the technology follow, or the reverse. The evidence from residential at scale points to the former. The counterargument deserves a hearing. ## Session details - **Session:** AI-Native Operations and Teams. Full title: AI-Native Operations: Redesigning Teams, Roles and Decision Workflows - **Track:** Artificial Intelligence - **Date and time:** Tuesday 22 September 2026, 11:30 to 12:20 CEST - **Location:** Innovation Auditorium 4, IFEMA Madrid - **Language:** English, with simultaneous translation into Spanish - **Access:** Premium VIP Pass - **Speakers:** Tripty Arya, Founder and CEO, Travtus. Anjli Carlucci, AI Engagement and Adoption Lead, Grosvenor Group - **Event:** The District 2026, 22 to 24 September 2026, IFEMA Madrid ## Frequently asked questions **What is AI-native operations?** AI-native operations is an approach in which a business is designed around AI being part of the working day, rather than adding AI tools on top of a process built for people alone. It changes three things in particular: who holds a decision, how quickly that decision can be reached, and who is accountable for the outcome. **What does it mean to be an AI-native real estate organization?** It means the operating model itself has been redesigned around AI, not only the tooling. The test is not how many AI products have been purchased. It is whether decision rights, approval paths and reporting lines have been re-cut to match the speed at which decisions can now be made, and whether the organization owns the intelligence it has built rather than licensing it. **How are real estate companies using AI in 2026?** The most widespread applications remain assistive: summarizing documents, drafting communications, surfacing exceptions in portfolio data, and answering questions that previously required manual investigation. More advanced work involves building organization-specific AI on a company's own operating data, so that outputs reflect that organization's standards and judgment rather than a vendor's defaults. **Why is the operating model, rather than the technology, the constraint?** Because AI changes the speed and confidence of a decision without changing the authority structure built around the previous speed. Where approval thresholds, escalation paths and role definitions still assume the old timescale, the organization acquires a capability it has no mechanism to use. **Is AI reducing headcount in real estate?** The available evidence does not support a confident directional claim, and outcomes vary substantially between organizations. What is consistently observable is a change in the composition of work: routine coordination is absorbed first, and the residual role is more judgment-dense. Planning for that change tends to be more productive than planning to a headcount target. **When and where is The District 2026?** The District 2026 runs from 22 to 24 September 2026 at IFEMA Madrid, having moved from Barcelona. It draws more than 15,000 attendees from across the international real estate investment market. --- *Travtus builds AI for housing. The observations in this article draw on the company's work with portfolios covering more than 500,000 homes across the UK and United States. Tripty Arya, Founder and CEO, speaks on AI-Native Operations and Teams at The District 2026 on 22 September.* ## What Does Your Second AI Use Case Cost? URL: https://www.travtus.com/resources/how-to-budget-for-an-ai-platform Date: 2026-09-03 Summary: Everybody in this market charges per door, so the rate isn't the question. What the door price buys is. How to build the budget for an AI platform, what actually comes off it, and the seven questions you'll get asked before anyone signs. > **Nothing.** That should be the answer. Right now you pay per door for leasing, then for collections, then for reporting, then for surveys, then for review management. You get the picture. The same doors, charged many times over. A platform is one cost, and whatever you build next sits inside it. Ask most operators what AI does for them today and you'll get a version of the same three things. A bot answering questions after hours and taking maintenance requests. Follow-ups chasing leads. Reminders on collections. All of it useful. And yes, you can build every one of those on Travtus, in a sentence, without waiting on a vendor to add it. But look at where that work sits. Every one of those jobs happens at the edge of your business, with a resident or a prospect on the other end. The bot takes the maintenance request. Then everything after it runs the way it always did. Somebody still prioritizes it. Somebody still finds a tech who isn't already on another job. **The perimeter gets faster. Nothing behind it changes.** Two things would change it, and neither one is on that list. The first is your own people getting answers without waiting on the data team. That changes who gets to make the call, not just how fast. The second is work that starts inside your business instead of with a resident. Between a regional and maintenance. Between asset management and operations. Nobody sells you that. Your own operations aren't a vendor's product. They're yours. That's the point, and it's also why this is hard to put in a budget. ## How do you build the budget? Start with the easy part. Everybody charges per door, so the rate isn't where this gets won or lost. What matters is that it's one cost for the platform, not a cost per function. When somebody asks for another workflow or another report next year, that isn't a new product, a new contract, or a new integration. It's the same line. Count how many times you're already paying per door on the same portfolio. One vendor for leasing, one for collections, one for reporting, one for procedures. That's the number your CFO hasn't seen laid out in one place, and it's the real shape of [AI vendor sprawl](/resources/reduce-ai-vendor-sprawl-real-estate). The hard part is what sits next to it. A brand new line with nothing above it looks like brand new money. And brand new money gets one question: **what comes out?** So have the answer ready before anybody asks. For most operators the list looks about the same: - **The data warehouse work**, and more to the point, what it costs to keep it fed and mapped. - **Reporting and analysis** you're paying for on the side. - **The small tools that each do one job** - pushing procedures out to site teams, sending somebody an alert when a number moves, reading resident messages for sentiment. - **The workarounds.** The spreadsheet that holds the real version of the number. The report somebody rebuilds by hand every month. The requests stacked up on your analyst's desk. That last one is the biggest item on the list and it isn't a vendor at all. You're already paying for it, out of payroll. It just never had a line of its own. Then phase it. Nobody funds the whole thing in year one. Give people access to the information first, and add the automation the year after. That halves the first ask, and it gives you a plan instead of a number. ## What does it replace, and what doesn't it? This is where most business cases fall over, so be blunt about it. Start with what actually comes off the budget. - **Mystery shopping.** You're paying somebody to sample a handful of calls a quarter and score them. Every conversation your team has is already being scored, all of them, every day. - **Your survey program.** Run the outreach here instead. And notice what you're really buying with a survey: you're asking a slice of residents once a quarter how they feel, when most of them have been telling you all year in ordinary messages. - **Replying to online reviews.** Generated and posted, without somebody on your marketing team writing them on a Friday afternoon. - **Chasing reviews.** Instead of blasting every resident with a review request, you ask the ones who just said something good, at the moment they said it. - **BI seats.** When regionals can ask their own questions, you need far fewer licenses for people who only ever opened a dashboard to find one number. - **The monthly reporting pack.** Built once, runs on a schedule, and arrives with the people who need it instead of being rebuilt by hand every month. - **Procedure and SOP tools.** Guides built from how your teams actually work, and kept current, rather than a library somebody updates once a year. - **Resident communication.** Inbound and outbound, on your rules, in your voice. Here's the part worth putting in front of your CFO: *every one of those is a separate vendor today, and most of them charge you per door. Same doors, over and over.* Now the other half of the list, and it matters just as much. These stay exactly where they are: - Your **property management system** - Your **leasing CRM** - **Rent pricing** - **Compliance training and certification** This sits on top of what you run and works with it. Nobody is moving your ledger, your leases, or your payments. Say that early, and say it to everybody. It costs you nothing and it does two jobs at once. It kills the worry that this is a system migration wearing a different hat, which is the first thing your IT lead will suspect. And it stops you promising savings you can't deliver, which is what gets a sponsor in trouble next summer when finance comes looking. **Only promise what you'll actually cancel.** A short honest list beats a long hopeful one. ## How do you make the case inside your own company? The number is the easy part. What you're really being asked is whether anybody will build anything with it. Here's how to have that answered before you walk in. 1. **Start with a decision somebody really makes.** Not a feature, a decision. Sit down with whoever owns the outcome and ask them to name one call they make regularly where better information would change what they do. That single sentence turns "a platform" into something with a dollar figure attached. It also makes the value theirs, not yours. 2. **Look at your own data before you pick the first project.** This feels backwards, and it's the strongest move you've got. A product demo turns into a feature comparison. Sitting down and looking at what you actually hold, and what shape it's in, turns into a conversation about what you could do with it. Then the first project gets picked against something real. 3. **Get other people to own it with you before you ask for money.** You want three people in this: you, somebody who owns the business outcome, and your sponsor. This isn't politics. It's the defense against the one objection that ends champions, and it's coming up next. If operations and asset management picked the first project, it's a business decision. If you picked it on your own, it's your bet. 4. **Pick the worst workaround, not the biggest opportunity.** The biggest opportunity has the most people involved and the longest road to proof. The worst workaround has a team that will use the thing on day one, and a before and after anybody can see. Maintenance coordination is usually sitting right there. 5. **Brief your sponsor in their own words.** Find the phrase they already use about this part of the business and hand it back to them. You're not translating your idea into their language. You're showing them something they already believe now has a way to happen. 6. **Ask for a commitment, not just a check.** Most people skip this and it's the one that holds the case together. Don't ask for approval to spend. Ask for approval to spend, *plus* a commitment from named teams to build a certain number of things in the first year. That last one matters more than it looks. A platform is poor value if you build two things and stop. Saying that yourself, before anyone else does, is what makes the rest of your case believable. ## What will you get asked? Seven questions. Knowing who asks each one is half the work. **"What comes out?"** Your CFO, and it'll be first. You've got the list. Keep it short and keep it real. **"We already tried this."** Usually somebody in IT or data, and this is the one that ends people. It's rarely an argument about the product. It's a reminder that money went into data once before and not much came back. You don't win it with a better answer. You win it by not being the only person in the room who wants this. If operations picked the project, the answer comes from them, not you. **"Why can't we just use Claude or ChatGPT?"** Often your CIO, and increasingly your CEO, because they've been using it themselves all year and it's genuinely good. Don't argue with them on that. Travtus runs on those models. This was never a choice between them. The question is what sits around the model, and there are three things a chat window doesn't have. - **It doesn't know your business**, and keeping it knowing your business is the actual work. A chat assistant answers from whatever you paste into it. Knowing what's going on at one community this week, across the work orders, the conversations, the lease and the ledger, kept current, is a data problem and not a prompting problem. - **Nothing starts on its own.** A chat window waits for somebody to open it. It won't notice a number moved on Tuesday and go tell the regional who owns that region. - **It's per person, not per company.** Everybody builds their own version in their own window. Nothing is versioned, tested, approved, published to a named audience, or scoped to a portfolio, and nobody can see what anybody else built. When they leave, it leaves with them. If your own team would rather wire a model up to your data themselves, that's a real option and worth pricing. Just go in clear-eyed that you're taking on the pipelines, the permissions and the housing know-how, and you'll own all three forever. **"Why not just buy a leasing AI?"** Nearly everybody asks some version of this, and it's fair. Three honest answers. First, a packaged set of workflows is a fixed set of somebody else's assumptions about your business. You can automate what they built and nothing else, and the work that's particular to how you run is out of scope for good. Second, none of it is internal, because your own operations aren't that vendor's product. Third, and this is the one to sit on: **the second thing you want to automate is another purchase.** On a platform, it's already paid for. That's the whole [platform versus point solution](/resources/platform-vs-point-solutions) question in one line. Worth mentioning quietly too, that small AI companies in this market have a habit of getting bought by bigger ones, and the product you signed for isn't always the product you end up with. **"Our PMS already has AI agents. Isn't it included?"** Bundled looks free. Four questions sort it out. 1. Who owns the logic you build? 2. Who owns the path your resident information travels, and how many companies does it pass through? 3. Who picks the model? 4. Who can change how it behaves - you, today, or somebody else's development list? Bundled agents also only reach as far as that vendor's own products, and they don't come with you if you ever leave. Those four belong on the shortlist in any [enterprise AI platform evaluation](/resources/how-to-evaluate-enterprise-ai-platform-real-estate), and the first one is really a question about [whose data model you're building on](/resources/build-ai-on-your-own-data-model). **"My teams won't use it."** Your COO, and they're right to ask. Worst workaround first. One team, one region. Make the first thing small enough that it can't fail quietly. **"Another vendor, another security review."** Procurement, legal, or IT, usually late and usually the thing that costs you a quarter. Fewer vendors is less to govern. Every extra tool is another agreement, another review, and another company handling resident information. Bring your counsel in early and briefly, rather than late and alarmed. ## How Travtus approaches this Travtus is the everyday AI platform for multifamily. Somebody on your team describes the workflow, report, or score they want, and they build it themselves, in a sentence. It stays yours. It sits on top of the systems you already run instead of replacing them, and it's built for the governance an enterprise actually needs. *A tool that does one job automates that job. A platform changes how the company runs.* The budget question is really just a question about which one you're buying, and that is most of what it means to be an [AI-native operator](/resources/what-it-means-to-be-an-ai-native-real-estate-operator). ## Before your 2027 budget locks We put together the one page a CFO will actually read: what the line looks like for your portfolio, what comes out to pay for it, and the questions you'll get asked with the names attached. ## Build AI on Your Data Model, Not Your Vendor's URL: https://www.travtus.com/resources/build-ai-on-your-own-data-model Date: 2026-08-11 Summary: The industry spent a decade building data foundations, then started buying AI inside point solutions — quietly handing the vendor the data model those foundations were built to protect. Here is what that trade costs, and how to keep your own model. > **Buying AI inside a point solution means adopting that vendor's data model — its entities, its hierarchy, its definitions.** Your own enterprise data model, the one your warehouse program spent years building, is what encodes how your business actually operates. AI should run on top of that model, not replace it. For most of the last decade, the serious data conversation in real estate was about foundations. Warehouse programs. Lakehouse migrations. Getting Yardi, RealPage, Entrata, work orders, surveys, and the general ledger into one place with one definition of a property, one definition of a portfolio, one definition of a resident. Slow, expensive, unglamorous work — and the right work. Then AI arrived, and a lot of that discipline went out the window. I spend my time with COOs and CIOs who have a dozen AI experiments running and no platform, and the pattern repeats: the leasing team bought an AI assistant, resident services bought a second one, the maintenance vendor shipped a third in a product update. Each of those tools is genuinely useful in its lane. And each one quietly asked the company to hand over the thing it spent years building — the model of its own business — in exchange for a feature. That is the trade nobody prices. Here is how to see it, and how to avoid it. ## Why buying AI from a point solution means adopting its data model Every application ships with an opinionated schema, and the AI inside it can only reason about the objects that schema defines. A leasing AI thinks in leads, tours, and conversations. A maintenance AI thinks in work orders, techs, and SLAs. A CRM thinks in contacts and stages. Those are reasonable primitives for the job each tool was built to do — but they are that vendor's primitives, designed for the average of their customer base. Your data model contains things theirs does not: the asset-level distinctions your investment committee actually uses, your fund and joint-venture structures, your regional groupings that do not match anyone's org chart, your renovation tiers, your definitions of a stabilized asset or an at-risk community, the operating standards that make your platform different from the operator down the street. That model *is* your operating model, written down. It is the reason two owners with identical unit counts perform differently. When AI lives inside a point solution, none of that exists. The AI is bright and completely uninformed about how your company sees the world. ## What you lose when the vendor's data model wins Three things, in escalating order of cost. 1. **Resolution.** Your entities get flattened into theirs. A question like "how are our value-add assets in their second renewal cycle performing against the stabilized set?" is not hard because the data is missing — it is hard because the tool has no concept of "value-add" or "second renewal cycle." Your distinctions vanish at the boundary. 2. **Compounding.** Every workflow you build inside a point tool is built on that tool's objects, so it cannot be reused anywhere else. Ten tools, ten sets of logic, none of them additive. Meanwhile the warehouse — the one asset that *would* compound — sits underneath, feeding dashboards. 3. **Optionality.** This is the expensive one. Once your workflows are expressed in a vendor's schema, your roadmap is their roadmap. You can only automate what they have decided to expose. And you cannot leave without rebuilding, which is exactly why the schema is designed that way. That is the handcuff. Not a contract term — an architecture. It is also how [AI vendor sprawl](/resources/reduce-ai-vendor-sprawl-real-estate) turns from an annoyance into a structural problem: each tool takes a piece of your operating model with it. ## Who owns the data in an AI platform? Ownership has three separate questions inside it, and vendors are usually only clear about the first one. - **Who owns the records?** Almost always you. This is the easy answer, and it is in every contract. - **Who owns the model — the entities, relationships, and definitions?** This is the one that matters, and it is frequently the vendor. If the platform can only represent your business using its own object types, you own the rows but not the meaning. - **Who owns the derived layer — the enrichment, the scores, the classifications, the conversation context the AI generates?** Ask explicitly. Ask whether it is queryable, exportable, and traceable back to a source record. If the intelligence your operation generates can only be viewed inside the vendor's interface, you are renting your own history. A short version to take into a vendor call: *if the platform cannot represent an entity that exists in your business but not in their product, they own the model, not you.* ## Build vs buy: do you have to build your own AI? No — and this is where the argument usually goes wrong. The instinct after this diagnosis is to build in-house, because building is the only way most teams have seen to keep control. But building a durable AI platform means owning orchestration, retrieval, evaluation, permissioning, model churn, and safety — permanently, with a team that competes for talent against every AI lab. Very few operators want that to be a core competency, and the ones who try usually end up with a good prototype and a stalled program. The real choice is not build vs buy. It is **which layer you buy at**. Buy at the point-solution layer and you buy a schema along with the feature. Buy at the [platform layer](/resources/platform-vs-point-solutions) and you buy infrastructure — orchestration, governance, security, integration — that maps onto the model you already own. Same procurement process, opposite outcome for optionality. The evaluation question that separates the two: *"Can we define a new entity, workflow, or metric that your product has never seen before — and can our business users do it without your engineering team?"* If the answer needs a roadmap conversation, you are buying at the wrong layer. It belongs on the shortlist of questions in any [enterprise AI platform evaluation](/resources/how-to-evaluate-enterprise-ai-platform-real-estate). ## How to unlock data trapped in your PMS and CRM You do not move it. You put a context layer above it. The systems of record stay exactly where they are — Yardi, RealPage, Entrata, AppFolio, your ticketing system, your survey tool, your warehouse. The platform layer connects to them and holds the thing none of them individually has: a unified model of your company, portfolios, communities, units, residents, and the operational events that connect them, expressed the way *your* business defines them. That is the difference between an AI that can answer "how many open work orders are there?" and one that can answer "which communities are showing maintenance and sentiment patterns that put renewals at risk next quarter?" The second question is not a harder model problem. It is a data-model problem — it requires a system that knows those objects are related in your business, because your model says so. Three practical rules if you are building this layer: - **Do not wait for perfect data.** Start with the systems you have access to and expand. Teams that hold the AI program until the warehouse is finished ship nothing for two years — the point [clean data alone was never going to solve](/resources/from-etl-to-intelligence-why-clean-data-wont-be-enough-in-the-age-of-ai). - **Keep the semantic definitions in your control**, not in a vendor's configuration you cannot export. - **Insist on lineage.** Every AI answer should be traceable to the source record it came from. Explainability is not only a compliance requirement — it is how the business learns to trust the output. ## How Travtus approaches this Travtus is built as the Everyday AI™ Platform — enterprise infrastructure that sits *above* your systems of record rather than replacing any of them. The platform integrates with the stack you already run and builds a context layer over it, so the AI reasons about your company the way your business defines it: your portfolios, your hierarchy, your operational reality. No [rip-and-replace](/resources/how-to-make-a-multifamily-company-ai-native), and no schema surrender. That is the deliberate design choice behind everything else. Because the context layer belongs to you, the workflows built on it belong to you too — they are not trapped inside one department's tool, and they compound instead of fragmenting. Explore, our conversational layer over that context, exists for the same reason: business users get answers from the model directly, without an analyst in the middle, and customers see around a 15% productivity improvement simply from making information easier to reach. Your data model stays yours. The AI is what you build on top of it. ## Frequently asked questions **Who owns the data in an AI platform?** You should own three things: the source records, the data model itself, and the derived intelligence the platform generates. Contracts usually cover the first and stay vague on the other two. Ask specifically whether enrichment, scores, and classifications are exportable and traceable to their source — if they are only visible inside the vendor's interface, you do not functionally own them. **Should we build our own AI platform or buy one?** For almost all operators, buy — but buy at the platform layer, not the point-solution layer. Building means permanently owning orchestration, evaluation, security, and model churn. Buying at the platform layer gets you that infrastructure while keeping your own data model, which is the part that actually differentiates your business. **How do we unlock data trapped in our PMS and CRM?** Add a context layer above the systems of record rather than migrating out of them. The platform connects to your existing systems and unifies them into one model of your business, so AI can reason across sources. You get cross-system answers without a rip-and-replace or a two-year migration. **Can an AI platform work with our existing data warehouse?** It should complement it, not compete with it. Your warehouse is the governed foundation; the AI platform is the layer that reasons and acts on top of it. Ask any vendor how their platform reads from your warehouse, whether derived outputs can flow back into it, and how lineage is preserved in both directions. **What is the difference between an AI platform and an AI point solution?** A point solution automates one task inside one department, using its own data model. A platform provides shared infrastructure — context, governance, orchestration — that multiple departments build on, using yours. The practical test: a point tool's workflows cannot be reused outside it, while a platform's compound across the business. *Want to see what this looks like on your stack? [Talk to Travtus](/demo) about where the context layer would sit — bring the systems you run today and the questions your team cannot currently answer.* ## AI for Housing: One Platform Across Every Rental Segment URL: https://www.travtus.com/resources/ai-for-housing Date: 2026-08-03 Summary: Housing is the one asset class where the asset talks back. Why AI for housing is its own category — not generic real estate AI — and how one platform serves multifamily, single-family rental, student, and affordable portfolios. > **AI for housing is artificial intelligence built for residential portfolios — multifamily, single-family rental, student, and affordable.** It's a distinct category from generic real estate AI for one structural reason: housing is the asset class where the asset talks back. Millions of resident conversations a month are the operating surface, and serving them takes one platform across every segment — not a tool per task. Ask what makes housing different from every other real estate asset class and you get answers about unit counts and lease terms. The real answer is simpler: **housing is the only asset class with people living inside it.** An office generates rent rolls and sensor data. A rental portfolio generates conversations — on the Travtus platform alone, millions of them every month, resolving into [more than 1,100 distinct types of request](/resources/what-renters-want-july-2026) in a single month. That's why "AI for housing" is not "AI for real estate, applied to apartments." It's its own discipline, and this piece is the map of it. ## What is AI for housing? AI for housing is artificial intelligence applied to the operations of residential portfolios: understanding and resolving resident communications, triaging maintenance, running move-ins and renewals, automating recurring workflows, and reading the operational signals that lead the financials. In its working 2026 form it's a platform layer — one governed system connected to the PMS and communication channels a housing business already runs. The residential specifics are what separate it from the broader [AI for real estate](/resources/ai-for-real-estate) picture: - **The volume is conversational.** The busiest surface isn't documents — it's people: requests, complaints, questions, notices, at a scale no team can staff for and no single-task bot can cover. - **The stakes are human.** A missed work order is someone's home. Handoff to a person — recognized, routed, with context — is a design requirement, not an edge case. - **The rules are real.** Fair-housing obligations mean consistency isn't just service quality — it's compliance. AI that applies the same neutral rules to every resident, with an audit trail, is structurally safer than ad-hoc variance. ## Does one AI platform work across multifamily, single-family, student, and affordable? Yes — and this is the argument for treating housing as one category rather than four products. The work is the same shape everywhere: residents communicating, maintenance arriving, leases turning, documents flowing. What changes is the mix and the rules, and that's configuration, not a new product. - **Multifamily** is the proving ground — the highest conversation density and the deepest deployments. (Our [multifamily AI guide](/resources/what-is-multifamily-ai) covers it in full.) - **Single-family rental** flips the geometry: scattered sites, no on-site staff, vendor-heavy maintenance. Centralized AI matters *more* when there's no office to walk into — the platform becomes the front desk for a portfolio spread across a metro. - **Student housing** compresses the entire lifecycle into an annual cycle: mass turns, guarantor-heavy applications, first-time renters with dictionary questions. Volume spikes that break staffing models are exactly what automation absorbs. - **Affordable housing** carries the heaviest documentation and compliance load per unit — recurring certifications, income documentation, audit trails. Document handling and status communication, done consistently and recorded, is where AI removes the most weight. One platform, four segment mixes. An operator running more than one segment — increasingly the norm among enterprise owners — gets the same context layer, the same workflows, and one security review across all of it. > **The catch:** buy AI segment by segment and task by task, and you rebuild the fragmentation the platform was meant to remove — a leasing bot here, an SFR maintenance tool there, an affordable-compliance system somewhere else, none of them sharing context. The segment differences are real, but they're the last mile, not the foundation. ## What should a housing AI platform do? Four tests, in order of how often vendors fail them: 1. **Connect to what's already there.** The PMS, the warehouse, the communication channels stay; the platform reads across them. No rip-and-replace. 2. **Understand residents end to end.** Not answer FAQs — resolve requests across channels, and hand off to a person with full context when judgment, emotion, or legal weight is involved. 3. **Let your teams build.** Workflows, scores, and reports described in plain language by the people who hold the judgment — [that mechanism in full](/resources/ai-workflows-housing) — so the logic stays the firm's IP. 4. **Carry governance in every feature.** Role-based security, property-level permissioning, audit trails — built in once, not bolted onto each tool. In housing, this is what makes AI deployable at all. Anything that passes one test in one segment is a point tool. The platform question — and why it decides everything downstream — is covered in [AI platform vs point solutions](/resources/platform-vs-point-solutions). ## How Travtus approaches AI for housing Housing-first is the whole design. Travtus is the [Everyday AI™ Platform](/everydayai) for housing — built on millions of real resident conversations, deployed across multifamily, single-family rental, and student portfolios, with enterprise governance in every feature. Operators report roughly 95% automation on the workflows they build and around 15% productivity gains from conversational access to their own data — and the context that powers one segment powers the next, because it's one platform. The category is young enough that most firms are still assembling it from parts. The ones that aren't — the ones that picked the housing platform first — are the ones whose operations now compound. ## Frequently asked questions **What is AI for housing?** Artificial intelligence built for residential portfolios — resident communications, maintenance, move-ins, renewals, workflows, and risk signals — running as one platform across multifamily, single-family rental, student, and affordable housing. **How is it different from AI for real estate generally?** Housing is the asset class where the asset talks back: the operating surface is millions of resident conversations, with human stakes and fair-housing obligations generic real estate AI isn't built for. **Does one platform serve all rental segments?** Yes — the work is the same shape everywhere; the segment differences are configuration, not separate products. **What should a housing AI platform do?** Connect to existing systems, understand residents end to end, let your teams build in plain language, and carry governance in every feature. **Is it fair-housing safe?** Designed properly, consistency at scale with audit trails is a fair-housing strength — neutral rules applied identically to every resident, with no role in screening. (Educational context, not legal advice.) --- *Running more than one segment? [See what one housing platform looks like](/platform), or [book a demo](/demo).* ## AI for Property Management: The 2026 Operator's Guide URL: https://www.travtus.com/resources/ai-for-property-management Date: 2026-08-03 Summary: What AI for property management actually does in 2026, how the leading companies deploy it day to day, and the one buying decision that determines whether any of it compounds. > **AI for property management is artificial intelligence running the everyday operations of rental housing — resident conversations, maintenance triage, recurring workflows, reporting, and risk signals.** In 2026 the divide isn't who has AI (89% of operators have introduced it) but who runs on it: the roughly one-third who embedded AI into daily operations did it on a platform, not a pile of point tools. Property management is a strange industry to automate: the work is enormous in volume, endlessly varied, and almost none of it is glamorous. That's exactly why AI has landed harder here than in most of real estate — and why the results have split so sharply between companies that bought AI tools and companies that changed how they operate. This guide is the map: what AI for property management actually does today, how the leading companies deploy it, what the data says, and the one buying decision that determines whether any of it compounds. ## What is AI for property management? AI for property management is artificial intelligence applied to the daily operations of rental housing at portfolio scale. In production today — not on a roadmap — that means five things: - **Resident conversations, handled end to end.** [AI agents](/resources/ai-agents-for-property-management) understand a request, take the action, and hand off to a person with full context when judgment is needed. - **Maintenance, triaged as it arrives.** A free-text complaint becomes a categorized, prioritized, correctly routed work order — and the day [re-sequences itself](/resources/maintenance-triage-multifamily) when an emergency lands. - **Workflows your own team authors.** Operations leaders describe how a situation should be handled in plain language, and [the platform builds the workflow](/resources/ai-workflows-housing) — no engineering ticket. - **Answers from your own operating data.** Instead of a report request, a manager asks "which properties had the biggest jump in open work orders this month?" and gets the answer with the data behind it. - **Signal before it becomes financials.** Every interaction carries sentiment, maintenance patterns, and retention risk that platform-level AI scores and aggregates while problems are still operational. The useful distinction in 2026 is not AI versus no AI — everything claims AI. It's *where the AI sits*: a feature inside one system, a standalone tool for one task, or a platform layer that works across the operation on shared context. That third form is where the results concentrate, for reasons the data below makes plain. ## How are property management companies using AI in 2026? The adoption story has two halves, and both matter. **The surge is real.** [34% of property management professionals used AI in 2025, up from 21% a year earlier](https://naahq.org/flat-rents-ai-adoption-2025-appfolio-property-management-benchmark-report-reveals-key-trends), and by mid-2026 [89% of multifamily operators have introduced AI in some form](https://www.globenewswire.com/news-release/2026/07/29/3335276/0/en/EliseAI-Unveils-New-State-of-AI-in-Multifamily-Report-85-See-Reduced-Operating-Expenses.html). The motivation is structural, not fashionable: onsite turnover runs at [29.2% a year](https://naahq.org/news/why-employee-retention-challenges-go-deeper-wages), and [two-thirds of property management leaders' time goes to routine and reactive work](https://www.appfolio.com/blog/naa-top-challenges-2025). The labor math forces the question. **But embedding is rare.** In that same 2026 survey, only about a third of operators have fully embedded AI into daily operations. The gap between "introduced" and "embedded" is the industry's real divide — and it tracks scale: [47% of operators managing more than 5,000 units use AI against 28% of the smallest](https://www.bisnow.com/national/news/top-talent/apartments-ai-driven-revolution-creating-tech-haves-and-havenots-128182). The embedded camp looks different on the ground: fewer vendors, one governed platform connected to the PMS they already run, and AI in the daily flow of work rather than in a side dashboard. We track this divide in depth in [the state of multifamily AI](/resources/state-of-multifamily-ai-2026). ## How can property managers use AI day to day? Here's what "embedded" actually looks like across a working week — because the value isn't in a monthly report, it's in the everyday: - **The inbox stops being triage.** Routine requests — amenity questions, account queries, document submissions — resolve the moment they arrive, on any channel. Our platform data shows why this matters: renter demand spans [more than 1,100 distinct request types in a single month](/resources/what-renters-want-july-2026), and no single type exceeds 6% of volume. No team can staff for that distribution; AI with context can serve it. - **Maintenance runs on your rules.** Intake, routing, reminders, and status updates happen automatically; the director's triage logic — what jumps the line, who gets what — runs as a workflow instead of living in one person's head. - **Questions get answered where they're asked.** A regional manager preparing for an owner call asks the platform directly instead of filing a report request. Teams using Explore this way see around a 15% productivity gain simply from easier access to information. - **The playbook runs itself.** "When a work order reopens a third time, flag the unit and tell the regional manager why" — described once, applied consistently at every property. Operators report roughly 95% automation on the specific workflows they build. - **People handle what needs people.** Distressed residents, legal weight, genuine exceptions — the point of automating the volume underneath is that humans are available for exactly these, with the full history attached. Start with [what to automate first](/resources/property-management-automation-what-to-automate-first) if you want the sequencing; the short version is high-volume, low-judgment work before anything clever. ## What results should you expect — and what's the catch? The claimable outcomes cluster in three places: workflow automation rates (roughly 95% on flows operators build), productivity (around 15% where conversational access replaces manual digging), and retention (improvements near 25% where service quality scales without headcount). The survey data rhymes: [firms implementing AI expect 31% portfolio growth versus 12% for firms without it](https://www.appfolio.com/newsroom/property-manager-benchmark-survey-2026). The full economics — including how to measure it honestly — are in [the ROI of AI in property management](/resources/roi-of-ai-in-property-management). > **The catch:** none of these results come from a shelf of disconnected tools. A leasing bot, a maintenance add-on, and a reporting copilot each see a fragment of the same resident and share nothing — which is why so many pilots stall at "introduced" and never reach "embedded." The results above belong to the platform camp. That's the one buying decision that matters. A point tool automates a task a vendor chose; a platform is infrastructure your own team builds on, with context across departments, one security review, and economics that compound — the second use case costs a fraction of the first. The full framework is in [AI platform vs point solutions](/resources/platform-vs-point-solutions), and the diligence questions are in [how to evaluate an enterprise AI platform](/resources/how-to-evaluate-enterprise-ai-platform-real-estate). ## How Travtus approaches AI for property management Travtus is the [Everyday AI™ Platform](/everydayai) for housing operators — the platform-camp answer. It connects to the PMS you already run (Yardi, RealPage, Entrata, AppFolio — no rip-and-replace), builds one operational picture from your communications and operating data, and puts your teams in the builder's seat: agents handling resident conversations end to end, workflows and reports authored in plain language, scores that surface risk early, and [data feeds](/resources/what-renters-want-july-2026) that keep your data yours. Start with one use case at a defined set of properties; expand on the same context. The companies pulling ahead in 2026 aren't the ones that bought the most AI. They're the ones that operate on it, every day. ## Frequently asked questions **What is AI for property management?** Artificial intelligence applied to daily rental-housing operations: resident conversations, maintenance triage, recurring workflows, self-service data answers, and early risk signals — delivered as single-task tools or, where the results concentrate, as a platform layer across the whole operation. **How are property management companies using AI in 2026?** Adoption surged (34% of professionals in 2025, up from 21%; 89% of multifamily operators have introduced it) but only about a third have embedded AI into daily operations — and the embedded camp runs it as a platform connected to their existing PMS. **How can property managers use AI day to day?** Routine requests resolve on arrival, maintenance triages itself on the team's own rules, playbooks run automatically at every property, and managers ask questions of their data in plain English instead of filing report requests. **What is the ROI of AI in property management?** Roughly 95% automation on workflows operators build, ~15% productivity gains from self-service data access, retention improvements near 25% — and firms with AI expect 31% portfolio growth versus 12% without. **Point tools or a platform?** Renter demand is a 1,100-type long tail no single-task tool can serve. Anchor on a platform with shared context; keep genuine specialists on top. --- *Ready to move from "introduced" to "embedded"? [See the platform](/platform), or [book a demo](/demo) and bring one use case.* ## AI for Real Estate: Where It's Actually Working in 2026 URL: https://www.travtus.com/resources/ai-for-real-estate Date: 2026-08-03 Summary: Real estate AI has moved from experiments to operating advantage — but unevenly. Where AI is actually working across the lifecycle in 2026, from acquisitions to operations, and what separates the firms compounding from the firms piloting. > **AI for real estate in 2026 is past the experiment phase: it prepares acquisitions from the broker inbox, flags portfolio risk before it reaches the financials, runs everyday operations, and answers questions from a firm's own data.** The divide is no longer who has AI — it's who embedded it: the firms compounding run one platform across the lifecycle; the firms still piloting bought tools. Real estate spent two decades as the economy's least-digitized major asset class, and then compressed a decade of technology adoption into about three years. The money has noticed: [$16.7 billion went into proptech in 2025, up 67.9% year over year](https://www.multifamilydive.com/news/proptech-investment-venture-capital-funding/809517/), with investors explicitly favoring AI built on strong data foundations. But capital chasing a category tells you where the market thinks value is — not where it actually is. Having built AI for housing since before the current wave, here's my honest map of where AI for real estate is genuinely working in 2026, and where it still stalls. ## What is AI for real estate? AI for real estate is artificial intelligence applied across the property lifecycle — not one tool for one task, but intelligence wherever decisions and work happen: 1. **Acquisitions.** Deals arrive as unstructured broker email; AI extracts the asset and terms, resolves duplicates, scores against the buyer's own criteria, and chases missing documents — so analyst time goes to judgment, not data entry. ([The full acquisition pipeline](/resources/multifamily-acquisition-pipeline).) 2. **Asset management and risk.** The risks that hurt you aren't in the financial statements yet. Operational signals — sentiment, maintenance patterns, renewal behavior — lead the financials by months, and AI reads them at portfolio scale. 3. **Operations.** The everyday work of running buildings: resident conversations handled end to end, maintenance triaged on arrival, recurring playbooks running as [workflows the team describes in plain language](/resources/ai-workflows-housing). This is the deepest deployment surface — our [property management guide](/resources/ai-for-property-management) covers it in full. 4. **Decision support.** Leadership asking questions of portfolio data directly — in plain English, with the underlying data attached — instead of waiting in a reporting backlog. Note what's *not* on this list: rent-setting algorithms. Pricing AI is under active litigation in residential real estate, and it's not what the working deployments are built on. The value is in operations and intelligence, not in pricing. ## How is AI used in real estate in 2026 — and how far along is the industry? The honest answer: further than the skeptics think, less far than the marketing says. [Deloitte's 2026 Commercial Real Estate Outlook](https://www.deloitte.com/us/en/insights/industry/financial-services/commercial-real-estate-outlook.html) — surveying 850+ executives at firms with $250M+ AUM — found only 19% of CRE executives still describe their organizations as early in the AI journey, but 27% report real implementation challenges: technical hurdles, missing expertise, internal resistance. In residential specifically, [89% of multifamily operators have introduced AI in some form, yet only about a third have fully embedded it into daily operations](https://www.globenewswire.com/news-release/2026/07/29/3335276/0/en/EliseAI-Unveils-New-State-of-AI-in-Multifamily-Report-85-See-Reduced-Operating-Expenses.html). That introduced-versus-embedded gap is the industry's real state of play, and it splits by scale: [47% of operators managing more than 5,000 units use AI against 28% of the smallest](https://www.bisnow.com/national/news/top-talent/apartments-ai-driven-revolution-creating-tech-haves-and-havenots-128182). The embedded firms share a pattern we've documented across [the state of multifamily AI](/resources/state-of-multifamily-ai-2026): one platform, connected to the systems they already run, with their own teams building on it. ## Is real estate AI actually delivering results? Where it's embedded — measurably. On platform deployments, operators report roughly 95% automation on the specific workflows they build, around 15% productivity gains where conversational access to data replaces manual digging, and retention improvements near 25% where service quality scales without headcount. The surveys rhyme: [85% of operators using AI report reduced operating expenses](https://www.globenewswire.com/news-release/2026/07/29/3335276/0/en/EliseAI-Unveils-New-State-of-AI-in-Multifamily-Report-85-See-Reduced-Operating-Expenses.html), and [firms implementing AI expect 31% portfolio growth versus 12% for those without](https://www.appfolio.com/newsroom/property-manager-benchmark-survey-2026). There's also a capital-markets dimension the operating metrics miss: **being AI-native is becoming part of how firms differentiate to raise money.** When returns are driven less by financing and more by how consistently you run the assets you own, "our operations are systematically better" is an investor story — and it's only credible when it's platform-deep, not pilot-thin. That argument in full: [what it means to be an AI-native real estate operator](/resources/what-it-means-to-be-an-ai-native-real-estate-operator). > **The catch:** the results concentrate almost entirely in the embedded camp. A deal-flow tool, a leasing bot, and a BI copilot bought separately each hold a fragment of the same asset's story and share nothing — which is why so many real estate AI initiatives produce a dozen pilots and no operating change. Fragmentation, not model quality, is why initiatives stall. ## What separates the firms compounding from the firms piloting? Three decisions, consistently: 1. **They bought a layer, not a shelf.** One governed platform with context across the lifecycle, instead of a point solution per department. The economics are the tell: on a platform, the second use case costs a fraction of the first; on a shelf, every use case costs like the first. ([Platform vs point solutions](/resources/platform-vs-point-solutions).) 2. **They connected rather than replaced.** The systems of record stayed — PMS, warehouse, communications — and the AI layer reads across them. No rip-and-replace program, which is where transformations go to die. 3. **Their own people build.** The logic that matters — what makes a deal interesting, what makes an asset at-risk, how a request should be handled — is described by the people who hold that judgment, in plain language, and it stays the firm's IP. ([How to evaluate an enterprise AI platform](/resources/how-to-evaluate-enterprise-ai-platform-real-estate) has the twelve diligence questions.) ## How Travtus approaches AI for real estate Travtus is the [Everyday AI™ Platform](/everydayai) for housing — built for exactly that embedded camp, across multifamily, single-family rental, and student housing. It connects to the stack a firm already runs, builds one operational picture from communications and operating data, and puts every lifecycle surface on it: deal context that survives closing, portfolio signals that lead the financials, agents and workflows the team authors, and answers on demand from the firm's own data. The firms pulling ahead in 2026 aren't running AI experiments. They're running their business on AI — and that, not the tool count, is what the market is starting to price. ## Frequently asked questions **What is AI for real estate?** Artificial intelligence applied across the property lifecycle — acquisitions, portfolio risk, operations, and decision support — delivered in its working form as one platform layer connected to a firm's existing systems. **How is AI used in real estate in 2026?** Acquisitions teams score deals from broker inboxes, asset managers read operational risk signals, operating teams automate conversations and maintenance, and leadership queries portfolio data in plain English. Most CRE executives are past early exploration; execution is the differentiator. **Is it delivering results?** In the embedded camp, yes: ~95% automation on built workflows, ~15% productivity gains, retention improvements near 25%, and 85% of AI-using operators reporting reduced operating expenses. **Why do initiatives stall?** Fragmentation — tools bought per department that share nothing. 89% of operators have introduced AI; only about a third embedded it. **Does AI mean replacing existing systems?** No. The working pattern is a platform connected to the systems of record, adding intelligence on top. --- *Where would your firm land on the introduced-versus-embedded line? [See the platform](/platform), or [book a demo](/demo).* ## AI Workflows in Housing: How to Build One by Describing It URL: https://www.travtus.com/resources/ai-workflows-housing Date: 2026-08-03 Summary: Building automation has always meant a specification, a backlog, and an engineering dependency — so the people who know what the workflow should do never build it. That constraint has changed: describe the workflow, and the platform builds it. Almost every housing business now has AI somewhere. Very few have AI doing anything that the business itself designed. The reason is not appetite. It is that building automation has always required a specification, a place in a development backlog, and somebody technical to do the work. The person who knows what the workflow should do has historically never been the person able to build it, and by the time the request reaches the front of the line the need has usually moved. That is the constraint worth paying attention to, because it is the one that has changed. ## What is an AI workflow? An AI workflow is a sequence of steps the software carries out on its own: noticing something, gathering what it needs, deciding what follows, and acting, without a person moving it along at each stage. That is different from a chatbot, which answers, and from a dashboard, which displays. As our explainer on [agentic AI](/resources/agentic-ai-multifamily-operators) puts it, the value "is not in any single action. It is in the ability to coordinate multiple actions intelligently." In housing, the useful examples are rarely exotic. They are the recurring sequences that currently depend on somebody remembering to do them. ## What does it mean to build one conversationally? It means you describe the outcome you want in ordinary language, and the platform assembles the workflow. No specification document, no ticket, no waiting. Describe *when a work order is reopened for the third time, flag the unit and tell the regional manager why*, and that is the build. Describe *every week, score each asset on the signals we care about and rank them worst to best*, and that is the build too. The significance is not the saved typing. It is that the description is written by the person who understands the judgment being encoded. Nothing is lost in translation to a developer, because there is no translation. ## Why does it matter what the workflow is connected to? Because a workflow is only as capable as the data it can reach. This is where most automation quietly fails. A tool that automates inside its own four walls can only act on what it already holds, so anything requiring two systems to agree stays manual. A workflow built on a platform runs across everything connected to it, which means one description can span a property management system, a communications record and a document store without anybody building an integration for that specific case. That is also why the connected layer matters more than the cleverness of any individual step. Without it, AI stays confined to isolated tools. ## What do these workflows look like across different roles? The same mechanism produces very different things depending on who describes it. **Acquisitions.** A deal arrives as an email. The workflow extracts the asset and the terms, recognises it as the same opportunity already sent by another intermediary, requests the missing rent-roll detail from the broker, and scores it against the buyer's own criteria before anyone opens it. See [the acquisition pipeline in full](/resources/multifamily-acquisition-pipeline). **Asset management.** Every asset carries a daily score built from the signals that particular strategy cares about, ranked worst to best, with the reason attached rather than the number alone. More on [real-time portfolio visibility](/resources/real-time-portfolio-visibility-multifamily). **Data and analytics.** The recurring reports that consumed a morning each week build themselves and arrive on a schedule, with each answer carrying its source. More on [taking ad-hoc requests off the data team](/resources/automated-report-generation-real-estate). **Resident service.** Routine requests are handled end to end, and anything carrying emotion, unusual complexity or legal consequence is handed to a person with the full conversation attached. More on [keeping tone human across channels](/resources/ai-resident-comms-that-dont-feel-like-a-bot). **Maintenance.** When an emergency lands or a technician calls in sick, the plan re-sequences itself according to the triage rules that team already applies in its head. None of these required an engineering programme, and none required replacing the system of record underneath. The [playbook for becoming AI-native without ripping out your stack](/resources/how-to-make-a-multifamily-company-ai-native) sets out how that sequencing works. ## Where do AI workflows not help? This is worth being straight about, because the answer tells you where to start. Workflows are strongest where work is high in volume and low in judgment: it recurs constantly and follows rules a trained new hire could apply. They are weakest where work is genuinely novel, relationship-heavy, or requires a person to be accountable for the outcome. A negotiation is not a workflow. Nor is a difficult conversation with a resident, nor a decision that has legal weight. The practical version: automate the volume underneath so people can spend their attention on the judgment on top. Teams building this way typically automate around 95% of the steps in the flows they build, leaving people to handle the exceptions, and see roughly a 15% productivity gain from the time no longer spent hunting for information. Human review is a design decision rather than an afterthought. Where a workflow touches residents, vendors or colleagues, the point at which a person takes over is part of what gets described, and fair, neutral language is the default rather than a review step. Workflows built on Travtus play no part in screening or in decisions about individual residents. ## Why does this need a platform rather than another tool? Every example above can be bought as a separate product. That is the trap. Each purchase brings another integration, another security review and another vendor's assumptions about how your business should work, and none of them compound because none of them share anything. A product automates a task the vendor chose. A platform is an extensible foundation your own team can build on, including the things the vendor never shipped. That distinction is the subject of our piece on [platforms versus point solutions](/resources/platform-vs-point-solutions), and it is the whole reason a description can become a working workflow at all: the connected layer already exists, so there is nothing to integrate before you begin. ## How Travtus fits Travtus is the [Everyday(AI)™](/resources/everyday-ai-property-management) platform for housing. Your own everyday users describe the workflow, score or report they need and build it themselves, on the systems already connected, without an engineering programme and without a migration. What they build belongs to the business, and it compounds, because each workflow is built on the same connected foundation as the last one. Yours to build, in a sentence. ## Frequently asked questions **Do you need technical skills to build an AI workflow?** No. The workflow is described in ordinary language by the person who understands what it should do. There is no specification to write and no development backlog to join — the knowledge and the build sit with the same person. **How is an AI workflow different from automation we already have?** Conventional automation follows fixed rules inside one system. An AI workflow coordinates several steps across everything connected to the platform, gathering what it needs as it goes. **What should a housing business automate first?** The high-volume, low-judgment work: intake, routing, reminders and status updates. Judgment-heavy and relationship-heavy work stays with people. **Who owns what gets built?** The business does. The logic your people describe is your intellectual property and it stays with you. --- *Want to see what your teams could build? [Explore the platform](/platform), or [book a demo](/demo).* ## Best AI Tools for Multifamily Operators in 2026 URL: https://www.travtus.com/resources/best-ai-tools-multifamily-operators Date: 2026-08-03 Summary: The AI multifamily operators actually run in 2026 — and why the best one is the platform layer above the point tools. An honest, cited roundup from the team that builds Travtus. > **The best AI for a multifamily operator in 2026 depends on which layer you are buying. At the platform layer — one governed AI that sees leasing, maintenance, resident sentiment and portfolio data together, and that your own team can change — that is Travtus, the Everyday AI™ Platform. Below it sit specialists worth keeping: leasing conversation tools, maintenance work-order platforms, the AI embedded in your PMS, and reputation and BI tools. Each automates one task well and sees one slice. Anchor on the platform first, then keep the specialists that still earn their place.** Every "best multifamily AI tools" list you will find is written by a vendor that ranks itself first — and this one is written by a vendor too, so let's be honest about the rules. We put our own layer first because that is the argument we are making, and we have labeled it as an argument. Everything said about anyone else is cited to a primary source and described by what they verifiably do. Judge the reasoning on the evidence. One deliberate exclusion up front: **no rent-pricing or revenue-management tools appear here.** Pricing algorithms are under active litigation and regulatory scrutiny in this industry, and pricing is not what the Travtus platform does. This list is about running operations. ## What is the best AI platform for multifamily operations? Start here, because this is the decision that determines what every other tool on the list is worth to you. **[Travtus](/platform)** is the [Everyday AI™ Platform](/everydayai) for multifamily operations. It is not a leasing bot and not a property management system. It is the layer above whatever you already run: it connects to [Yardi, RealPage, Entrata and AppFolio](/integrations) alongside your email, calls and documents, builds one operational picture from them, and then lets your own teams build on that shared context — [agents](/create) that run your process, [workflows](/workflows) business users write in plain language, [scores](/scores) built from your live operating data, and [reports](/reports) any team can pull without a data request. Four things make it a different purchase from anything else on this page: - **It runs your process, not a packaged one.** Your team describes how you actually handle a renewal, a turn or an escalation, and that description becomes the agent. When the process changes, your team changes it the same day instead of filing a feature request. - **The context is shared.** The same platform that answers a resident can see the maintenance history, the sentiment trend and the renewal date, because they are not in four different vendors' systems. - **The logic and the data stay yours.** Travtus is model-agnostic and sits on your systems. What you build is portable if your stack changes. - **Governance and traceability are built in**, not bolted on per vendor — so when compliance asks why the AI said what it said, there is an answer. Cortland, MAA, BH and Continental build their own AI on Travtus. Operators on the platform report roughly 95% automation on the specific workflows they build. **[Book a demo](/demo)** and bring your tool list. > **The honest limit:** Travtus will not be your system of record, and it will not price your rent. If that is what you are shopping for, this is the wrong list. Four questions worth asking of every vendor below — including us: *Who owns the logic you build? Who owns the data? Can you see why it did that? Who controls the roadmap?* ## What AI tools do multifamily operators use for leasing and resident communication? The leasing conversation lane is the most crowded in multifamily, and the tools in it are good at what they do. **EliseAI** is the best-known specialist — automated leasing conversations, tour scheduling and resident communication across text, email, chat and voice, now with a CRM and maintenance app alongside it. **Funnel** takes a CRM-first approach built for centralized leasing: one guest card per renter across the portfolio, with AI handling 24/7 inquiry response and follow-up. **BetterBot** focuses on automated prospect engagement — FAQs, tour booking, follow-up — with multilingual, fair-housing-conscious response design. If your leasing process already looks like the one a packaged agent ships with, you get years of refinement without having to describe anything. What you give up is the ability to change it when your process moves. > **Where Travtus fits:** the same conversations across the same channels, but running the process your team describes rather than the vendor's — and on a platform where the leasing agent can see the maintenance history that explains why last year's residents did not renew. Conversation automation without operational context answers fast, but not always right. Detail: [Travtus vs EliseAI](/compare/eliseai) and [Travtus vs Funnel](/compare/funnel). ## What AI tools handle multifamily maintenance? **HappyCo** is multifamily-first and runs an [AI-powered centralized maintenance platform](https://www.businesswire.com/news/home/20240409748746/en/HappyCo-Unveils-AI-Powered-Centralized-Maintenance-Platform), with its JoyAI companion handling scheduling, technician matching and 24/7 resident maintenance comms. **Property Meld** covers the full work-order lifecycle and [acquired Mezo in January 2025](https://www.prnewswire.com/news-releases/property-meld-acquires-mezo-advancing-ai-driven-property-maintenance-operations-302350847.html) to add AI intake and triage. **Lula** pairs AI triage with a vetted technician network across roughly 60 US markets, troubleshooting the resident's report, assigning urgency, dispatching and tracking to completion. These are real specialists, and the technician network in particular is not something a platform replaces. > **Where Travtus fits:** Travtus does not dispatch trades or run a vendor network — keep the tool that does. What it does is stop the work order being an isolated ticket. The three-month leak and the lease expiring in sixty days are the same story, and the platform is the only place both are visible. [See how workflows automate the work around the ticket](/workflows). ## What AI is built into the property management systems? All the major PMS vendors now ship AI, and if you run their suite you should use it: - **[Entrata Layered Intelligence (ELI)](https://www.prnewswire.com/news-releases/multifamily-leader-introduces-entrata-layered-intelligence-infusing-ai-across-its-platform-302067268.html)** spans leasing, maintenance triage and — via its [2024 acquisition of Colleen AI](https://www.entrata.com/press/entrata-acquires-colleen-ai) — collections and renewals. - **[AppFolio Realm-X](https://www.appfolio.com/newsroom/appfolio-unleashes-realm-x-ai-capabilities)** embeds a conversational interface and AI agents for leasing and maintenance in AppFolio Property Manager. - **Yardi Virtuoso** offers AI agents across maintenance, invoicing and compliance, plus [connectors that let LLMs query live Yardi data](https://www.yardi.com/blog/introducing-yardi-virtuoso/). - **RealPage Lumina**, [announced in June 2025](https://www.realpage.com/news/realpage-showcases-lumina-ai-at-apartmentalize-2025/) in collaboration with OpenAI, adds an agentic AI workforce to the RealPage/Knock stack. > **Where Travtus fits:** embedded AI is the definition of single-system context — each reasons over its own suite's data and stops at the suite's edge. That is fine for in-suite tasks and it cannot give you one view across a portfolio running mixed systems, which most enterprise portfolios do. Travtus is not a PMS and does not compete for the ledger; these are [integrations](/integrations) first and comparisons second. Detail: [Travtus vs Entrata](/compare/entrata) and [Travtus vs RealPage](/compare/realpage). ## What about reputation, analytics, lease audit and smart-home AI? Four categories the big roundups mostly skip: - **Reputation:** J Turner Research is multifamily-only — its ORA® score is the industry's online-reputation benchmark, and its [AI model was trained on 1.4 million apartment reviews](https://www.multifamilyexecutive.com/property-management/j-turners-ai-model-analyzes-1-4-million-reviews_o) across 22 categories of the living experience. Opiniion (resident feedback and review automation) and Widewail (review generation and managed response) round out the category. - **Lease audit and due diligence:** SurfaceAI is purpose-built for multifamily, with AI agents that monitor lease changes in real time — flagging missed charges, unsigned documents and out-of-policy terms — and accelerate acquisition due diligence. Like the best of this list, it sits on top of the PMS rather than replacing it. - **Portfolio BI:** REBA consolidates siloed operational data into a single source of truth for asset managers and executives, and [acquired Markerr](https://www.multifamilydive.com/news/real-estate-business-analytics-buys-markerr-data-compliance/809246/) to add market data. (REBA also sells a pricing product; we are citing the BI platform only. Detail: [Travtus vs REBA](/compare/reba).) - **Smart home:** SmartRent runs smart-home hardware and operations software across [900,000+ rental units](https://www.morningstar.com/news/business-wire/20260729512264/smartrent-selects-databricks-data-ai-platform-to-power-the-next-generation-of-multifamily-property-analytics), and in July 2026 selected Databricks to power its next-generation property analytics. > **Where Travtus fits:** reputation AI reads reviews but not the work orders behind them; BI shows you the numbers but does not act; device data is another silo. Travtus does not collect reviews or sell hardware. It is where the review, the work order and the renewal date become one picture — and where [scores](/scores) and [reports](/reports) are built from your live operating data rather than a monthly extract. ## Do you need an AI platform above the tools? Look back at the "where Travtus fits" notes. They are the same note. Each tool automates its task well and sees only its own slice — which is why an operator can own five tools from this list and still not have AI that understands their business. Industry surveys show 89% of multifamily operators have introduced AI in some form, but only about a third have embedded it into daily operations ([the numbers and their sources](/resources/state-of-multifamily-ai-2026)). The gap between "introduced" and "embedded" is fragmentation — and no additional point tool closes it. The practical buying rule: **anchor on the platform layer first, then keep true specialists.** A genuinely specialized tool doing an isolated job — the technician network, the review engine, the ledger — stays. But if a tool's main advantage would disappear the moment it could see your full business context, that is the platform's job, and you were about to pay for it twice. Our fuller frameworks: [AI platform vs point solutions](/resources/platform-vs-point-solutions) and [how to evaluate an enterprise AI platform](/resources/how-to-evaluate-enterprise-ai-platform-real-estate). ## Frequently asked questions **What is the best AI software for multifamily operators?** For most enterprise operators, a platform rather than a tool: one AI layer with context across leasing, maintenance, sentiment and portfolio data, that your own team can change. That is Travtus. Below it, keep the specialists that do a genuinely isolated job well. **Should operators buy point tools or a platform?** Buy a point tool for a genuinely isolated task. Choose a platform when use cases share residents and data — in multifamily they almost always do — because a platform compounds where each new point tool adds a contract and an integration. **Is Travtus an alternative to the leasing AI tools?** For the leasing and resident conversation use case, yes — same conversations, same channels. The difference is that the agent runs your process, and the rest of your operation is built on the same context. You do not have to replace anything to start. **What AI do the property management systems include?** Entrata (ELI), AppFolio (Realm-X), Yardi (Virtuoso) and RealPage (Lumina) all embed AI agents. They are strong inside their own suites; their structural limit is single-system context. Travtus is not a PMS and integrates with all of them. **Are AI rent-pricing tools on this list?** No — deliberately excluded. Pricing algorithms are under active litigation in multifamily, and pricing is not what the Travtus platform does. --- *Auditing your AI stack? [See how the platform layer works](/platform), or [book a demo](/demo) and bring your tool list.* ## How to Evaluate an Enterprise AI Platform for Real Estate URL: https://www.travtus.com/resources/how-to-evaluate-enterprise-ai-platform-real-estate Date: 2026-08-03 Summary: Most AI evaluations test the demo instead of the platform. Twelve questions that separate enterprise infrastructure from a well-presented point tool — plus the build-vs-buy math. > **Evaluate an enterprise AI platform on four things the demo won't show: whether it reasons across your whole business or one department's slice, whether your own business users can build on it without engineering, whether governance is in every capability rather than bolted on, and whether it integrates without replacing your system of record.** Everything else is presentation. I've been on both sides of these evaluations, and the pattern that decides most of them is unhelpful: the vendor with the best demo wins. That's understandable. A demo is concrete and a platform architecture is abstract. But a demo shows you one path, on the vendor's data, tuned by someone who knows exactly where the edges are. It tells you almost nothing about whether the thing will still be earning its keep in eighteen months when you want it to do something nobody scripted. So here's the diligence I'd actually run, organized around the four questions that separate infrastructure from a well-presented point tool. Then the build-vs-buy math, which is usually the real question underneath. ## What's the difference between an AI platform and an AI point solution? The marketing distinction is meaningless — everything calls itself a platform. The functional distinction is **extensibility**. A point solution automates a defined task, within one department, using the data it can reach. It does that task well. When you want it to do a different task, you file a feature request and wait for a roadmap. A platform holds context across the business and lets you build many capabilities on it, including ones the vendor never anticipated. The test is simple and worth running in the evaluation itself: *ask them to build something they didn't plan for, using your data, in front of you.* Not a customization of an existing template — something new. How that goes tells you more than the rest of the process combined. If the answer involves a services engagement, a professional-services quote, or a roadmap conversation, you're buying a point solution with good positioning. ## What questions should you ask an AI vendor in real estate? Twelve, grouped. I'd want defensible answers to all of them before signing. **On context and data** 1. What data does the platform reason across — one department's records, or communications *and* transactional data *and* documents together? Fragmented context is the ceiling on everything else. 2. What do you actually need from our data team to get started, in hours? "Just an API key" and "a six-month pipeline project" are both answers, and only one is acceptable. 3. What happens if our data is incomplete or messy in places? An honest vendor has a real answer here, because everyone's data is. **On who can build** 4. Can a business user — an operations manager, not an engineer — create a new capability unaided? Ask to watch someone non-technical do it. 5. Who owns the logic and IP our teams create on the platform? Get this in writing. 6. What's the path when we want something you haven't built? See the extensibility test above. **On governance and trust** 7. How does permissioning work at the property and portfolio level? In housing this is non-negotiable — the wrong person seeing the wrong asset's data is a real problem, not a theoretical one. 8. Is governance built into every capability, or applied per-integration? Bolted-on governance fails at the seams, and the seams are where the incidents happen. 9. Is there an approval step before something a team builds goes live, and is there an audit trail of who changed what? This is what makes internal adoption defensible to your compliance function. 10. Are you tied to a single model provider? Model-agnostic architecture is what stops today's procurement decision from becoming tomorrow's lock-in. **On the commercial and operational shape** 11. Can we start on one use case or one property and expand, without a platform-wide commitment up front? 12. What does the second capability cost, in time and money, relative to the first? On a real platform it should be dramatically cheaper. If it's the same, each capability is really a separate product. Question 12 is the one I'd weight most heavily. Point-tool economics are linear or worse — every addition brings its own integration burden. Platform economics compound. That difference is the entire financial case, and it shows up in the answer to a single question. ## Should you build or buy an AI platform for property management? For virtually every housing operator, the answer is **buy the platform, build your own capabilities on it** — and it's worth being precise about why, because "build vs buy" is usually framed too coarsely. Building the *platform* means committing, permanently, to: - Data infrastructure and integration engineering across a shifting vendor landscape. - Model operations — evaluation, versioning, cost management, keeping current as models change every few months. - Security, governance, and permissioning engineering to enterprise standard. - The talent market for all of the above, competing against technology companies. None of that differentiates you. No owner-operator has ever won a deal because their internal model-ops practice was excellent. You'd be funding a permanent cost center to reach parity. Building your own *capabilities* — the scores, the workflows, the reports that encode how your business actually operates — is entirely different. That **is** your differentiation. Your definition of renewal risk, your escalation logic, your view of asset health: those are competitive advantages precisely because they're yours and nobody can buy them. So the question isn't build or buy. It's **which layer you build at.** Buy the infrastructure; build the logic. An operator that buys a point solution has bought someone else's logic, which is the worst of both — you're paying for it and it isn't yours. ## How long should an enterprise AI platform evaluation take? Weeks, structured around one real capability rather than months of demos. A defensible pilot: pick one use case with a clear owner and a measurable outcome. Scope it to one property or one department. Connect real data. Build the capability. Then — and this is the part most pilots skip — have **your own team** build the second one, with the vendor watching rather than driving. That last step is the whole evaluation. If your operations manager can build the second capability, you have a platform and the economics compound. If it requires the vendor, you have a services relationship with a product attached, and every future capability will cost what the first one did. ## How Travtus approaches this I'll answer the twelve honestly rather than restate them as features. Travtus is the [Everyday AI™ Platform](/everydayai) for housing, and the design intent is enterprise infrastructure rather than a departmental tool. On context, it reasons across communications, records and numbers, and the documents and outputs published to it — that combination is the point, and it's what a single-department tool can't reach. On who builds: your business users describe what they want in a sentence and build it in Studio, which is why the second capability costs a fraction of the first. On governance, data governance, role-based security and property-based permissioning are built into every feature rather than applied per-integration, workflows carry an approval step before they publish, and it's model-agnostic by design. On integration, it works alongside [Yardi](/integrations/yardi), [Entrata](/integrations/entrata) and the rest of your stack — no rip-and-replace, and you can start with one use case or one property and expand. The honest limitation to note in any evaluation: this is a platform, which means it rewards operators who want to build their own logic and is a poor fit for anyone looking for a single turnkey automation and nothing else. If you want one narrow job done and never intend to extend it, a point tool is a reasonable purchase. Related reading: [AI platform vs point solutions](/resources/platform-vs-point-solutions) on the category distinction, [how to reduce AI vendor sprawl](/resources/reduce-ai-vendor-sprawl-real-estate) if you're consolidating, and [is AI for property management secure and enterprise-ready?](/resources/secure-enterprise-ready-ai-property-management) for the security diligence in depth. Plans and packaging are on [Plans](/plans); the security detail is on [Security](/security). ## Frequently asked questions **How do you evaluate an enterprise AI platform for real estate?** Test four things the demo won't show you: whether it reasons across your whole business or just one department's data, whether your own business users can build on it without engineering, whether governance and permissioning are built into every capability rather than bolted on, and whether it integrates without replacing your system of record. **Should you build or buy an AI platform for property management?** Buying the platform and building your own capabilities on top is usually the right answer for housing operators. Building the platform itself means funding data infrastructure, model operations, governance, and security engineering indefinitely — none of which differentiates you. Your differentiation is in the logic your teams build, which is why that part should stay yours. **What questions should you ask an AI vendor in real estate?** Ask what happens when you want a capability the vendor didn't build, who owns the logic and IP your teams create, how permissioning works at the property level, whether the platform is tied to one model provider, and what the integration actually requires from your data team. Vague answers on any of these are informative. **What's the difference between an AI platform and an AI point solution?** A point solution automates a task within one department using the data it can see. A platform holds context across the business and lets you build many capabilities on it. The practical test is extensibility: if you can't create something the vendor didn't anticipate, you have a point solution regardless of how it's marketed. --- *Run the twelve questions against us. [Start with the platform architecture](/platform), or [book a demo](/demo).* ## Maintenance Triage in Multifamily: How to Re-Sequence the Day When It Breaks URL: https://www.travtus.com/resources/maintenance-triage-multifamily Date: 2026-08-03 Summary: Every maintenance director builds a plan in the morning, and every maintenance director knows it will not survive the morning. How to encode the triage rules you already carry — and let the day re-sequence itself. Every maintenance director builds a plan in the morning, and every maintenance director knows it will not survive the morning. An emergency lands. A technician calls in sick. Parts do not arrive. By ten o'clock the plan is a document describing a day that is no longer happening. What follows is the part nobody schedules: somebody re-sequences the whole thing by hand, and that somebody is almost always the person who can least afford the time. ## Why does the plan break so reliably? Because the work does not arrive at a steady rate, and the roster does. [Our analysis of more than 21,000 resident and prospect conversations](/resources/are-multifamily-properties-understaffed) found resident interaction running about two and a half times higher in the first and last two days of each month, with more than a third of it arriving outside ten to six. Staffing is level against a workload that is not. A plan built for an average day is wrong on the days that matter most. The usual response is to ask for more people. The industry benchmark already sits at roughly one extra full-time employee per 45 apartments, and adding to it treats a distribution problem as a capacity problem. ## What are triage rules, and where do they live? Every operation has them, and almost no operation has written them down. They are the working model of what jumps the line: which unit, which resident, which technician, what waits until tomorrow, what escalates immediately regardless. They account for building quirks, vendor lead times, who is trusted with what, and which resident had the same problem three weeks ago. That model is the most valuable thing in the operation and the least documented. Which is precisely why it has to be applied in person, all day, by the one person holding it. It does not scale, it does not survive their annual leave, and it leaves with them. ## Why is a scheduling module not the answer? Because a scheduling tool applies the rules its vendor chose, and those rules were written for a generic property. Most scheduling functionality can sort, assign and notify. What it cannot do is re-sequence a whole day against the specific logic of one operation, because that logic was never available to it. So the tool produces a plan, the day breaks it, and a person fixes it by hand. The tool is not wrong; it is simply solving the easier half. ## What should a maintenance operation automate first? The high-volume, low-judgment work: the tasks that happen constantly and follow rules a trained new hire could apply. - **Intake.** Turning a free-text complaint into a categorised, prioritised work order. - **Routing.** Getting each item to the right technician without a person deciding. - **Reminders and follow-ups.** The work that slips because nobody was watching the clock. - **Status updates.** Telling the resident where their request stands, before they ask again. None of it requires judgment and all of it consumes hours. Teams building these flows automate roughly 95% of the steps in them, leaving people to handle the exceptions. It is also worth knowing how much of the inbound is not maintenance at all. In that same analysis, about one in seven interactions was a maintenance request, while two in five were property and policy questions spread across more than 700 subtopics. A meaningful share of what reaches a maintenance team was never maintenance work. ## Where does a person have to stay in the loop? Anywhere judgment or risk is involved, and that line should be drawn deliberately rather than discovered. An emergency affecting habitability, anything with legal or financial consequence, and any situation where a resident is distressed all belong with a person, however routine the underlying request looks. The point of automating the volume underneath is that the people are available for exactly these. The same analysis found about one in four conversations escalating to someone on site despite a predictable outcome. That is the opposite failure: a person spending a trip on something the system could have resolved. ## How Travtus fits Travtus is the [Everyday(AI)™](/resources/everyday-ai-property-management) platform for housing. For a maintenance operation, the useful part is that the rules stay yours. Your team describes how the plan should re-prioritise when new work lands, in plain language, and builds the workflow that does it. The rules are the ones the director already carries. The re-sequencing happens as work arrives rather than at the next planning meeting, and it happens without anybody rebuilding the day by hand at four in the afternoon. Because it runs on a platform rather than inside a scheduling module, the same logic reaches the systems you already use, and changing it is a sentence rather than a change request. More on why that distinction matters in [platforms versus point solutions](/resources/platform-vs-point-solutions), and on the mechanism itself in [how to build an AI workflow by describing it](/resources/ai-workflows-housing). Yours to build, in a sentence. ## Frequently asked questions **Can work orders be re-prioritised automatically when an emergency comes in?** Yes. The workflow re-sequences the plan as new work lands, applying the operation's own triage rules rather than a vendor's defaults. What jumps the line, who it goes to and what waits are decisions the team defines and can change. **How is this different from the scheduling module in a work-order system?** A scheduling module applies rules chosen by its vendor and generally sorts, assigns and notifies. Re-sequencing an entire day against one operation's specific logic requires that logic to be available to the system — which is what a platform provides. **What should a maintenance team automate first?** Intake, routing, reminders and status updates: high in volume, low in judgment, quick to return hours. Judgment-heavy and resident-sensitive work stays with people. **Does automating triage remove people from the decision?** No. Emergencies, legal or financial consequence, and distressed residents belong with a person — automating the volume underneath is what makes people available for those. --- *Want to see your own triage rules running themselves? [Explore the platform](/platform), or [book a demo](/demo).* ## How to Build Your Multifamily AI Tech Stack URL: https://www.travtus.com/resources/multifamily-ai-tech-stack Date: 2026-08-03 Summary: A layered blueprint for the multifamily AI tech stack: what stays (your PMS), what's optional (task tools), and the platform layer that decides whether any of it compounds. > **Build a multifamily AI tech stack in three layers: keep your systems of record (the PMS), put a governed AI platform on top of them as the context layer, and add specialist tools only for genuinely isolated jobs.** Sequence matters more than selection: connect the platform first, prove one use case, then expand — the tool-first path is how operators end up with nine vendors and no compounding. Two-thirds of a property management leader's week goes to routine and reactive work — 42% routine operational, 24% firefighting — against just 16% on strategic work, according to the [AppFolio and National Apartment Association 2025 Performance Ecosystem Report](https://www.appfolio.com/blog/naa-top-challenges-2025). Onsite turnover is running at [29.2% a year](https://naahq.org/news/why-employee-retention-challenges-go-deeper-wages). That's the problem an AI stack exists to solve — and why building it in the right order matters more than picking the perfect logo at each position. Here is the blueprint we walk enterprise operators through, layer by layer. ## What should a multifamily AI tech stack look like? Three layers, each with a different buying rule. **Layer 1 — Systems of record. Keep them.** Your PMS ([Yardi](/integrations/yardi), RealPage, [Entrata](/integrations/entrata), AppFolio) and accounting stack manage leases, financials, and compliance. They stay. Every credible AI strategy in this industry is built *on top of* the systems of record, not instead of them. If a vendor's plan starts with replacing your PMS, that's a multi-year program wearing an AI costume. **Layer 2 — The AI platform. This is the decision that matters.** One governed layer that connects to Layer 1, builds context across departments — residents, properties, communications, work orders, documents — and lets your own teams create on it: agents that handle conversations, [workflows](/workflows) that encode your playbooks, [scores](/scores) that surface risk, [reports](/reports) that run themselves. This layer determines whether everything else compounds or fragments. It's also where governance should live: one security review, one audit surface, one place that controls what AI can do. **Layer 3 — Specialists. Optional, and fewer than you think.** Tools with deep domain machinery for genuinely isolated jobs — smart-home hardware, specialized inspection tooling. The test from our [tools roundup](/resources/best-ai-tools-multifamily-operators): if the tool's advantage would disappear the moment it could see your full business context, it belongs to Layer 2, not Layer 3. Most operators built this upside down: a stack of Layer 3 tools, no Layer 2, and then a discovery that nothing connects. That's the nine-vendor estate we described in [how to reduce AI vendor sprawl](/resources/reduce-ai-vendor-sprawl-real-estate). ## What order should you build the stack in? Sequence is the difference between compounding and sprawl. 1. **Connect the platform to what you already run.** No data-transformation project first — the platform brings the context layer and reads from the systems you have. This step disrupts nobody. 2. **Prove one high-volume use case.** Usually resident communications: it's where the labor is, and where the signal is. Scope it to a defined property set with a measurable outcome. 3. **Expand sideways on the same context.** Maintenance triage, retention scoring, self-service reporting — each new use case is cheaper than the last because the context already exists. This is the compounding that tool-by-tool stacks never reach. 4. **Audit Layer 3 annually.** As platform context deepens, some specialists stop earning their seat. Retire them at renewal — no cutover events. The data says the gap this sequence creates is real. [AppFolio's 2026 Benchmark Report](https://www.appfolio.com/newsroom/property-manager-benchmark-survey-2026) found firms implementing AI expect 31% portfolio growth versus 12% for firms without it — and adoption already splits sharply by scale, with [47% of operators managing more than 5,000 units using AI against 28% of the smallest operators](https://www.bisnow.com/national/news/top-talent/apartments-ai-driven-revolution-creating-tech-haves-and-havenots-128182). The enterprises are pulling away. ## What questions should you ask before adding anything to the stack? Five, before any AI purchase at any layer: 1. **Which layer is this?** If the vendor can't answer cleanly, it's a Layer 3 tool with Layer 2 marketing. 2. **What context does it reason over?** Its own silo, one suite's data, or your whole operation? 3. **Who builds on it — your team or their roadmap?** The stack should make your operations team more capable, not more dependent. 4. **What does the second use case cost?** Platform economics compound; tool economics repeat. (The full twelve-question version is in [how to evaluate an enterprise AI platform](/resources/how-to-evaluate-enterprise-ai-platform-real-estate).) 5. **Does it add a security surface or consolidate one?** Every additional vendor touching resident data is another review, another audit, another risk. ## How Travtus fits the stack Travtus is Layer 2 — the [Everyday AI™ Platform](/everydayai) built to be the context and creation layer of a housing operator's stack. It connects to the systems of record you already run, builds one operational picture from communications and operating data, and puts your teams in the builder's seat: describe the workflow, the score, or the report you want, and it runs on shared, governed context across the portfolio. Operators on the platform report roughly 95% automation on the workflows they build and around 15% productivity gains where [Explore](/explore) replaces manual digging. The result is a stack that gets simpler as it gets more capable — one platform deep in the middle, systems of record untouched beneath it, and only the specialists that truly earn their seat above it. ## Frequently asked questions **What should a multifamily AI tech stack look like?** Three layers: systems of record (keep), an AI platform with cross-departmental context (the core decision), and a small set of true specialists (optional). Most failed stacks were built tools-first with no platform layer. **Do I need to replace my PMS?** No — the PMS stays as the system of record; the platform connects to it. No rip-and-replace. **What order should adoption follow?** Connect the platform, prove one high-volume use case, expand sideways on shared context, audit specialists annually. **How many AI vendors should an operator have?** One platform for the connected core plus a few true specialists — not the nine-vendor estates common today. **Why do AI stacks fail?** Fragmentation: tools bought department-by-department hold context in silos and never compound. --- *Want the blueprint mapped to your actual stack? [See the platform](/platform), or [book a demo](/demo).* ## How to Reduce AI Vendor Sprawl in Real Estate URL: https://www.travtus.com/resources/reduce-ai-vendor-sprawl-real-estate Date: 2026-08-03 Summary: Nine AI vendors, each with its own integration, its own contract, and its own view of the resident. Consolidating point solutions into one platform is now an architecture decision, not a procurement one. > **You reduce AI vendor sprawl by consolidating the layer beneath the tools, not by swapping tools one for one.** Each point solution carries duplicate infrastructure, its own integration, and its own partial view of the resident. Moving that layer onto one platform removes the duplication and lets you retire individual tools gradually, at renewal, without a cutover. A CIO at a large owner-operator walked me through their AI estate last year. Nine vendors. A leasing assistant, a maintenance triage tool, a sentiment product, two separate reporting layers, a renewals tool, a call summarizer, and two things nobody could fully account for because the department that bought them had reorganized. Every one of those had been a defensible purchase. A department had a problem, a vendor solved that problem, the pilot went fine, it got renewed. Nothing irrational happened at any single decision point. The aggregate was a mess. Nine integrations to maintain. Nine contracts. Nine copies of resident data, each slightly stale relative to the others. And — the part that actually hurt — nine partial views of the same resident, none of which could see the others. When a resident who had a three-month maintenance problem got a cheerful renewal message, that wasn't a bug in the renewals tool. It was working exactly as designed, on the only data it had. ## What is AI vendor sprawl and why does it cost so much? Vendor sprawl is what you get when AI is bought department by department, the way software has always been bought. Each tool is narrow by design and each is sold on the strength of solving one team's problem well. The cost lands in four places, and only the first one shows up in procurement: **License fees.** The visible cost, and usually the smallest of the four. **Duplicated infrastructure.** This is the structural one. Every point solution bundles its own plumbing — its own email sending, its own telephony, its own data connections, its own storage. You are paying several times for the same underlying capability because it's embedded inside each application rather than shared beneath them. **The integration tax.** Making nine tools interoperate is an endless demand for APIs, webhooks, and vendor coordination. Each new tool multiplies the number of connections rather than adding one. This work never finishes, and it consumes exactly the engineering capacity you wanted to free up. **Data team drag.** Every vendor needs data, so every vendor generates a pipeline request. The data team ends up as an internal integration service for third-party products instead of building anything. There's a fifth cost that's harder to quantify and probably the largest: **no shared context**. Nine tools each holding a fragment means no tool can reason across the business, which is where the real value of AI sits. You've bought nine narrow automations and none of the intelligence. ## Should you consolidate AI tools or your property management system? Not the PMS. This is worth stating clearly because the two questions get conflated and it makes the consolidation decision look far scarier than it is. Your property management system is the system of record for leases and financials. It should stay. Your data warehouse, if you have one, stores and structures data — it should stay too. Neither is what's sprawling. What's sprawling is the **intelligence and automation layer above them**: the AI capabilities that read your data, decide something, and act. That layer is where you have nine vendors, and it's the layer that can consolidate without anyone touching the system of record. Getting this distinction right matters because it changes who has to approve the project. Replacing a PMS is a multi-year, board-level program. Consolidating the AI layer above it is neither. ## How do you consolidate point solutions without a rip-and-replace? The path that works in practice is gradual and runs alongside what you already have. **1. Connect first, replace nothing.** Stand up the platform and connect it to your existing systems — the PMS, the warehouse, communications channels. At this stage nothing is being retired and no team is being disrupted. You now have one place with context across the business, which none of your nine tools had. **2. Rebuild one capability.** Pick a single function currently owned by a point tool and build it on the platform. Something with a clear owner and a measurable outcome. Run both in parallel and compare. **3. Retire at renewal, not at cutover.** Once a capability works on the platform, let the corresponding tool lapse at its contract date. This is the key mechanic: there is no cutover event, no big-bang migration, no weekend where everything moves. Each renewal date is an independent, low-risk decision. **4. Repeat, and let the compounding do the work.** The second capability is easier than the first because the context is already there. The fifth is much easier than the first. This is the opposite of point-tool economics, where each addition increases integration burden. The reason this ordering matters: it means you're never betting the estate on a migration. If capability three doesn't work out on the platform, you keep the incumbent tool and move on. The downside at each step is one renewal cycle. ## Which AI tools are worth keeping? Consolidation isn't a maximalist exercise. Some tools should survive. **Keep** anything that owns a genuinely specialized function with real domain machinery behind it — where the vendor's value is deep expertise in a narrow, hard problem rather than a thin layer over your own data. Keep anything embedded in a regulated or contractual process where the disruption cost exceeds the consolidation benefit. And keep systems of record. **Consolidate** the tools whose value proposition is mostly "we put an AI interface on data you already own." That's the category where you're paying for duplicated infrastructure and getting a fragmentary view in return. In most estates I've looked at, it's the majority of the AI spend. A useful test: if the tool's main advantage would disappear the moment it had access to your full business context, it's a consolidation candidate. If it would still be better than anything you'd build, keep it. ## How Travtus approaches this Travtus is the [Everyday AI™ Platform](/everydayai) for housing, and the architectural claim is specific: it's the layer *above* point tools, not another one beside them. **[Connect](/integrations/yardi)** is what makes the gradual path possible. Travtus works with the systems you already run — [Yardi](/integrations/yardi), [Entrata](/integrations/entrata), and the rest of your stack — so there is no rip-and-replace to get started. The platform doesn't replace the PMS or the warehouse; it makes what they hold usable across teams. Then the capabilities get built on top, by your own people rather than by a vendor's roadmap: [Reports](/reports), [Profiles](/profiles), [Scores](/scores) and [Workflows](/workflows), each described in a sentence and built in Studio. That's what replaces the point tools one at a time — operators report roughly 95% automation on the specific workflows they build — and because they all sit on the same context, capability three benefits from what capabilities one and two established. The infrastructure to run this safely at enterprise scale — governance, security, property-based permissioning — is in the platform rather than bolted onto each tool, which is precisely the duplication you were paying for nine times. It's also model-agnostic by design, so consolidating onto the platform isn't a bet on one model vendor. The companion piece on the category logic is [AI platform vs point solutions](/resources/platform-vs-point-solutions), and if you're newer to the category, [What is multifamily AI?](/resources/what-is-multifamily-ai) sets the foundations. More in [Everyday AI & the AI-Native Operator](/resources/category/everyday-ai). ## Frequently asked questions **How do you reduce AI vendor sprawl in real estate?** Consolidate the layer beneath the tools rather than swapping tools one for one. Sprawl is expensive because every point solution carries its own infrastructure, integration, and data copy. Moving to one platform that holds the context and lets you build capabilities on top removes the duplication, and you can retire individual tools gradually as their function is rebuilt. **What is AI vendor sprawl and why does it cost so much?** It's the accumulation of narrow AI tools each bought to solve one department's problem. The cost isn't only license fees — it's the duplicated infrastructure inside each product, the integration and webhook work to make them interoperate, and the data team time spent moving the same records into several systems that never share context. **Should you consolidate AI tools or your property management system?** Neither replaces the other. Your PMS is the system of record for leases and financials and should stay. What consolidates is the intelligence and automation layer above it — the AI capabilities currently spread across several vendors. Consolidating that layer doesn't require touching the PMS. **How do you consolidate point solutions without a rip-and-replace?** Run the platform alongside what you have, connect it to your existing systems, and rebuild capabilities one at a time. Each capability that works on the platform lets you retire a tool at its renewal date rather than all at once. There's no cutover event, which is what makes the path viable for an enterprise. --- *Nine vendors, one resident, no shared view. [See what consolidating the layer beneath them looks like](/platform), or [book a demo](/demo).* ## What Is Multifamily AI? A Guide for Enterprise Operators URL: https://www.travtus.com/resources/what-is-multifamily-ai Date: 2026-08-03 Summary: Multifamily AI has come to mean everything from a leasing chatbot to enterprise infrastructure. Here is a plain definition, what it can actually do today, and the difference between owning AI tools and operating as an AI-native company. > **Multifamily AI is the use of artificial intelligence to run the everyday operations of apartment portfolios — resident communications, leasing, maintenance, reporting, and portfolio oversight.** Done well, it isn't a collection of chatbots. It's a platform layer that works across departments on shared context, so the whole company operates differently. I spend most of my weeks with COOs and CIOs who have already "done AI." They have a leasing bot at some properties, an AI in the call center, a pilot in collections. What they don't have is a company that operates differently than it did three years ago. That gap — between owning AI tools and being an AI-native operator — is what this guide is about. ## What is multifamily AI? Multifamily AI is artificial intelligence applied to the operations of rental housing at portfolio scale: understanding and responding to residents, automating recurring workflows, answering operational questions from your own data, and surfacing risk before it reaches the financials. The useful distinction is not "AI vs. no AI" — nearly every vendor now claims AI. The distinction is **where the AI sits**: 1. **Feature-level AI** — an AI capability inside a system you already own (your PMS adds a chatbot). 2. **Point-solution AI** — a standalone tool that automates one task or one department, such as leasing inquiries. 3. **Platform-level AI** — an intelligence layer that connects to your existing systems, builds context across departments, and lets your teams automate and ask questions anywhere in the business. Most operators start at levels one and two. The operators pulling ahead are the ones treating AI as infrastructure — level three — because context is what makes AI genuinely useful. An agent that can see a resident's full history, the property's maintenance load, and the team's process guides gives a different answer than a bot that only sees one inbox. ## What can AI actually do in multifamily operations today? Concretely, and in production today — not on a roadmap: - **Handle resident conversations.** [AI agents](/resources/ai-agents-for-property-management) resolve routine requests across channels — maintenance triage, amenity questions, account queries — and hand off to a human when judgment is needed. - **Automate recurring workflows.** Operations teams describe how a situation should be handled, and agents follow that playbook consistently at every property. Operators building these workflows on a platform report roughly 95% automation on the specific workflows they build — and on the Travtus platform, operators author those workflows themselves, no engineering ticket required. - **Answer questions from your own operating data.** Instead of waiting on a report request, a regional manager asks in plain English — "which properties had the biggest jump in open work orders this month?" — and gets an answer with the underlying data. Travtus customers using Explore, our conversational agent, see around a 15% productivity improvement simply from making information easier to reach. - **Turn conversations into intelligence.** Every resident interaction carries signal: sentiment, emerging maintenance patterns, retention risk. Platform-level AI scores and aggregates that signal so leaders see problems while they're still operational, not after they're financial. - **Generate and schedule reporting.** Insight reports that used to consume analyst time get generated from operating data, versioned, and re-run on a schedule. What multifamily AI does **not** credibly do is replace judgment. The best deployments put AI on the routine 80% so people can spend their time on the exceptions that actually need a human. ## What's the difference between AI tools and a multifamily AI platform? A tool automates a task. A platform changes how the company operates. | | Point AI tools | AI platform | |---|---|---| | Scope | One task or department | Cross-departmental | | Context | Sees only its own channel | Shared context across systems and teams | | Ownership | Each vendor a separate contract, security review, integration | One governed layer | | Who builds on it | The vendor's roadmap | Your business teams | | Compounding value | Flat — each tool is a ceiling | Rising — every workflow adds context for the next | The practical test I give executives: **when a new use case appears, do you buy another vendor, or configure it on what you already have?** If every new problem means a new procurement cycle, you have tools. If your operations team can stand up the new workflow on existing infrastructure, you have a platform. I've written a fuller comparison in [AI platform vs point solutions](/resources/platform-vs-point-solutions). ## How are enterprise multifamily operators using AI in 2026? Adoption has split by scale: large enterprise operators are well ahead of smaller ones, and their deployments increasingly look like consolidation plays — fewer vendors, one intelligence layer, AI in the daily flow of work rather than in a side dashboard. (For the fuller market picture, see [the state of multifamily AI in 2026](/resources/state-of-multifamily-ai-2026).) The pattern among REITs and PE-backed owner-operators we work with looks like this: 1. **Start with one high-volume use case** — usually resident communications, because that's where the labor and the signal both live. 2. **Connect the platform to the existing stack** — Yardi, RealPage, Entrata, AppFolio — rather than replacing anything. 3. **Expand sideways** — the same context that powers resident comms powers maintenance triage, retention signals, and self-service reporting, so each new use case gets cheaper to add. 4. **Make it everyday** — value shows up when on-site teams, regionals, and asset managers all touch the platform as part of normal work, not when a pilot report gets circulated. ## How do you get started without ripping out your stack? You don't need a data transformation project first, and you don't need to change your PMS. The platform approach works precisely because it sits *on top of* the systems of record you already run and brings the context layer with it. Start with one use case at a defined set of properties, prove the outcome, then expand — the platform grows with you. The mistake to avoid is the opposite path: accumulating single-purpose tools for two years and then discovering you've built vendor sprawl instead of capability. ## How Travtus approaches multifamily AI Travtus is the Everyday AI™ Platform for housing operators — the one AI platform an enterprise uses every day to become AI-native. It connects to your existing stack, builds deep operational context from your communications and operating data, and puts that context to work everywhere: AI agents your teams configure themselves, conversational access to your data through Explore, operational scores that surface risk early, and reporting that runs itself. The point isn't another tool. It's that the market — and your investors — can see you operate in a way competitors can't easily copy. ## Frequently asked questions **What is multifamily AI?** Multifamily AI is artificial intelligence applied to apartment portfolio operations — resident communications, leasing support, maintenance workflows, reporting, and portfolio risk visibility. It ranges from single-task tools to platform-level infrastructure that works across departments on shared context. **What is the difference between an AI tool and an AI platform for multifamily?** A tool automates one task in one department and sees only its own channel. A platform connects to your existing systems, shares context across departments, and lets your own teams build new workflows on it — so each use case makes the next one cheaper. **Does adopting multifamily AI mean replacing my property management system?** No. A platform approach integrates with the systems you already run — Yardi, RealPage, Entrata, AppFolio — and adds an intelligence layer on top. Operators typically start with one use case at a defined set of properties and expand from there. **How do multifamily operators measure ROI from AI?** Through operational outcomes: automation rates on routine workflows (operators report roughly 95% on the specific workflows they build), faster response times, productivity gains from self-service data access — around 15% with conversational access to operating data — and earlier visibility into retention and maintenance risk. **Is multifamily AI only for large portfolios?** Adoption is currently strongest among enterprise operators — REITs and PE-backed owner-operators — but the underlying model applies across housing, including single-family rental, student, and affordable portfolios. *Ready to see what an AI platform looks like on your portfolio? [Talk to Travtus](/demo) about starting with one use case.* ## What It Means to Be an AI-Native Real Estate Operator URL: https://www.travtus.com/resources/what-it-means-to-be-an-ai-native-real-estate-operator Date: 2026-08-03 Summary: Being AI-native isn't owning a pile of AI tools — it's an operating posture where AI is part of how the company runs every day. Here's what separates the two, and how operators get there. > **Being an AI-native real estate operator means AI is part of how the company runs every day — embedded across decisions and everyday work — not a collection of separate tools bolted onto individual teams.** It's an operating posture, not a tool count. Ten disconnected AI point solutions don't make you AI-native; one platform with context across departments does. Most operators I talk to already have AI in the building. A leasing bot here, a maintenance triage tool there, a review-response app the marketing team found. What they don't have is an answer to the question their board is starting to ask: *are we actually an AI-native business, or do we just own a lot of AI?* It's a fair question, and the distinction is not semantic. It's the difference between a company that has changed how it operates and one that has spent money on experiments that never left the department they started in. This is a strategy piece, not a product tour — but by the end you'll have a clean way to tell which side of the line you're on. ## What does it mean to be an AI-native real estate operator? An AI-native operator uses AI *every day*, across the organization, as part of how the work actually gets done — not as a set of side tools a few teams log into. The test is simple: if you removed the AI, would the way the company operates change, or would a few tasks just get slower? AI-native means the former. That's an operating posture, and postures are built, not purchased. You don't become AI-native by adding a fifteenth tool. You become AI-native when a single platform has enough context about your operation to act across leasing, maintenance, resident services, and reporting — so intelligence shows up in the flow of everyday work rather than in a dashboard nobody opens. ## What's the difference between using AI tools and being AI-native? A point tool automates a task. A platform changes how the company operates. That one line is the whole distinction. Point tools are seductive because they're easy to buy and easy to pilot. But they share a fatal limit: each one only knows about its own slice. The leasing bot doesn't know what maintenance sees; the review tool doesn't know what the resident said last week. You end up with a dozen narrow automations and no compounding — a state better described as *AI-fragmented* than AI-native. Being AI-native inverts that. Because the platform carries context across departments, the value compounds: the same understanding of a resident, a property, or a portfolio informs every interaction. That cross-departmental context is the thing point solutions structurally cannot give you, no matter how many you own. ## How does a multifamily company become AI-native? Not with a two-year transformation program, and not by ripping out your stack. The path that actually works is deliberately unglamorous: 1. **Start with a context layer, not a rip-and-replace.** Connect the systems you already run — Yardi, RealPage, Entrata, AppFolio. You don't need perfect data access to begin; that requirement is exactly what kills most transformations before they start. 2. **Pick one use case or one property.** Prove value where it's measurable. Nothing builds internal belief like a real result on real operations. 3. **Expand across the operation.** Let the platform compound. Each new use case is cheaper than the last because the context is already there. Done this way, "AI-native" stops being a slogan and becomes a series of concrete steps — the same argument I made in [platform vs point solutions](/resources/platform-vs-point-solutions), applied to the whole company rather than one decision. ## Why is operations becoming the investment strategy in real estate? Here's the part that makes this a boardroom conversation, not an IT one. For years, returns in this industry were driven mostly by financing and pricing. Increasingly, the differentiator is *operations* — how consistently and efficiently you run the assets you already own. When that's true, the operators who can run leaner and more consistently at scale win, and that capability is increasingly an AI capability. That reframes the pitch to capital, too. "Our operations are a competitive advantage" is a much stronger story to investors when you can show it's systematic and AI-native, not heroic and manual. Being AI-native is becoming part of how serious operators differentiate themselves to raise and deploy capital. ## How Travtus approaches this This is the whole reason Travtus exists as the [Everyday AI™ Platform](/everydayai) rather than another point tool. We're enterprise infrastructure — the platform with the context to act across departments, so AI shows up in the everyday work instead of in a report. It integrates with your existing stack, so you can adopt without changing everything, and it grows from one use case to the whole operation. If you want the plain-language version of the everyday side, start with [what Everyday AI in property management actually is](/resources/everyday-ai-property-management), then see [the platform](/platform). ## Frequently asked questions **What does it mean to be an AI-native real estate operator?** AI is part of how the company operates every day, across departments — an operating posture, not a count of tools. **What's the difference between using AI tools and being AI-native?** A point tool automates one task in one department; being AI-native changes how the whole company works, because a platform with context acts across departments. **How does a company become AI-native?** Start with a context layer over your existing systems, prove one use case, then expand — no rip-and-replace. **Do we need perfect data first?** No — a context-layer platform works with the systems and data you already have. **Is it only for large operators?** The posture applies at any size, but the payoff scales with structure; enterprise operators feel it most. --- *Want to see what AI-native looks like on your stack? [Explore the Everyday AI Platform](/everydayai), or [book a demo](/demo).* ## What Renters Want: July 2026 URL: https://www.travtus.com/resources/what-renters-want-july-2026 Date: 2026-08-03 Summary: Millions of renter conversations in July resolved into more than 1,100 distinct types of request — and no single type topped 6% of the volume. The first edition of What Renters Want, our monthly read on resident demand. > **In July 2026, the Travtus platform analyzed millions of renter conversations — and they resolved into more than 1,100 distinct types of request. No single type exceeded 6% of conversation volume, the ten most common together covered barely a fifth of it, and leasing — the part most AI tools automate — was only about 20% of the conversation, at the very peak of leasing season. Out of season it's lower still. Renter demand is a long tail, and a long tail is a platform problem.** *What Renters Want is our monthly report on what renters actually ask their landlords. The Travtus AI platform analyzes millions of renter conversations every month across enterprise portfolios; each edition distills them into the shape of resident demand. This is the first edition.* Most of the industry's picture of "what AI does in multifamily" comes from the leasing funnel, because that's where the first wave of AI lived. We sit somewhere different: across the whole operation, in the flow of every conversation renters have with their property managers. Ask what renters actually want, at that scale, and the answer is surprising — not because of what tops the list, but because of how little the top of the list matters. ## What do renters ask their landlords most often? The ten most common requests in July, as a share of all renter conversation volume: | # | What renters asked | Share of volume | | --- | --- | --- | | 1 | "Here's what I'm looking for" — search criteria | 5.7% | | 2 | Submitting a document | 2.4% | | 3 | Create a maintenance request | 2.2% | | 4 | Schedule my move-in | 1.9% | | 5 | Give notice to vacate | 1.5% | | 6 | What's the move-in policy? | 1.4% | | 7 | Request a call back | 1.3% | | 8 | Set up a tour | 1.3% | | 9 | Confirming a document was received | 1.2% | | 10 | "I'll send my documents shortly" | 1.2% | Read that table again: the single most common thing renters asked for is under 6% of the conversation, and the entire top ten adds up to about 20%. **Four-fifths of renter demand lives below the top ten**, spread across more than 1,100 further request types. There is no "main thing" renters want. There are a thousand things. ## How much of renter demand is actually about leasing? About a fifth. Add up everything related to finding, touring, and applying for a home — search criteria, tours, applications, income proof, promotions, availability — and it comes to roughly 20% of July's conversation volume. And remember what month this is: July is peak leasing season, when tours and applications hit their annual high. This ~20% is the leasing share at its **high-water mark** — out of season it falls well below that. Hold that against the industry's AI conversation, because the first generation of multifamily AI was built almost entirely for this fifth, sized to its busiest month. The other 80% is operations: move-in logistics (scheduling, policy questions, insurance, utilities — one of the largest post-funnel clusters), documents, maintenance, payments and disputes, renewals and transfers, amenity questions, and community life. Three July details worth pausing on: 1. **Renters gave notice to vacate more than three times as often as they asked to renew.** Departure announces itself; loyalty doesn't. If you're waiting for the renewal conversation to manage retention, you're managing it too late — the signal lives in the everyday interactions months earlier. 2. **Renters asked about the move-in policy more often than they asked for a tour.** The industry has spent five years building AI for the tour request; the bigger conversation was the one right after the lease is signed. And the honesty runs deeper: thousands of renters proactively told their landlord a payment would be late — before missing it. Give people an easy channel and they bring you the operational truth early. The question is whether anything on your side is listening. 3. **Renters insurance is a quiet giant — and half the story is pushback.** One administrative topic generated fourteen distinct request types and roughly 2% of all conversation volume: renters submitting policy details, asking what coverage their lease requires, checking whether their policy complies — and, in one of July's more pointed signals, disputing renters-insurance charges they believed were wrong. Each of those threads either gets resolved on the spot, policy checked and ledger corrected, or it becomes twenty minutes at somebody's desk. > **The catch for leasing-only AI:** a leasing bot sees the 20% and none of the rest — not the maintenance frustration that predicts the notice to vacate, not the move-in confusion that sets the tone for the tenancy, and none of the thousand smaller requests where service is actually won or lost. ## What does the long tail of renter requests look like? This is our favorite part of the data, and the part no point solution can touch. Nearly **half of July's request types appeared fewer than a hundred times each** — individually rare, collectively constant. A sample from the tail: - "Do the BBQ pits have a fee?" - "How do I book the volleyball courts?" - "What's the fee for the yoga room?" - "What does *Pullman kitchen* mean?" - "Can you set up ramps for moving day?" - "Is there a chess club?" And the single most unexpected thing in July's data: renters asked the platform to explain the industry's own vocabulary in **58 distinct flavors of "what does that mean?"** — *what does guarantor mean, what does processing fee mean, what does Section 8 mean, what does e-money order mean,* even *what does two-bedroom mean.* Nobody designs a feature for "be the dictionary." But when the channel actually answers, renters treat it like a knowledgeable friend — and every jargon question answered plainly is a small deposit of trust no leasing bot was scoped to make. Every one of these is a real renter with a real need, forming a real opinion about their landlord based on the answer. And here's the operational truth: **no software vendor will ever ship a feature for the BBQ pits.** You can't buy a point solution for the yoga room. The long tail can only be served by AI that answers from *your* operational context — your policies, your amenities, your process guides — and that's a platform capability by definition. The head of the curve tells you what to automate first. The tail tells you what kind of AI to buy. ## Why renter demand favors a platform over point solutions Put the three findings together — no dominant request, leasing at a fifth, a thousand-type tail — and the buying implication is hard to avoid. If demand were concentrated in a few big categories, a shelf of point tools could plausibly cover it. It isn't. Demand is a distribution, and the only thing that serves a distribution is a platform that can automate *anything on it*: the high-volume workflows your teams formalize, and the long-tail questions answered from shared operational context — with [roughly 95% automation on the workflows operators build](/resources/state-of-multifamily-ai-2026). That's the argument we make in [AI platform vs point solutions](/resources/platform-vs-point-solutions) and [what is multifamily AI](/resources/what-is-multifamily-ai), and this data is why we keep making it: the shape of what renters want *is* the case for the platform. ## How Travtus approaches this This report exists because the [Everyday AI™ Platform](/everydayai) handles these conversations end to end — understanding each request, acting on it, and handing off to a human with context when that's the request. The head of the curve runs as [workflows](/workflows) your teams author; the long tail is answered from your own operational knowledge; and every interaction feeds the [scores and signals](/scores) that tell you what's changing before it reaches the financials. And the data you've been reading about isn't ours to hoard — it's yours. Every client can configure their own **data feeds** out of the platform: self-service, scheduled exports of the data they choose, filtered how they want, delivered straight to their own warehouse and BI tools with no engineering request. Your renters' conversations become your data asset, not a vendor's. We'll publish What Renters Want every month. If you want to know what your renters are asking — and how much of it could be handled the moment they ask — [book a demo](/demo). ## Frequently asked questions **What do renters ask property managers most often?** In July 2026 the top request — describing the home they're searching for — was only 5.7% of conversation volume, followed by document submission (2.4%), maintenance requests (2.2%), and move-in scheduling (1.9%). Millions of conversations resolved into over 1,100 distinct types of request — breadth only an AI platform can serve. **How much of renter communication is about leasing?** Roughly a fifth of July's conversation volume — and July is peak leasing season, so that's the annual high-water mark. The other ~80% was operations — which is why the full renter lifecycle needs an AI platform, not a leasing-only tool. **Is multifamily AI just a leasing chatbot?** The data says it can't be: over 1,100 distinct request types in one month, none above 6% of volume. Serving that distribution requires an AI platform that can automate anything on shared context. **Can AI handle rare or unusual renter requests?** Only platform AI can — nearly half of July's request types appeared fewer than a hundred times each. They're answered from your own operational context, not a point-solution vendor's feature list. **Can operators access the data behind their renter conversations?** On Travtus, yes — every client can configure their own data feeds out of the platform: scheduled, filtered exports to their own warehouse and tools, no engineering request. Point solutions typically keep that data locked inside the tool. **What is the What Renters Want report?** The monthly report from Travtus, the Everyday AI™ Platform, on what renters actually ask their landlords — distilled from the millions of conversations the platform analyzes each month. --- *Next edition: August 2026. Want to see what your renters are asking? [Book a demo](/demo).* ## Multifamily Acquisition Pipeline: From Broker Email to Asset Handover URL: https://www.travtus.com/resources/multifamily-acquisition-pipeline Date: 2026-07-31 Summary: Deal flow arrives as unstructured, duplicated email — and everything learned in diligence is abandoned at closing. How to run an acquisition pipeline that starts at arrival and survives the handover. Acquisition teams are rarely short of deals. They are short of time to prepare them, and the reason is unglamorous: the pipeline starts in an inbox. Broker notes, teasers, forwarded threads, a rent roll as a PDF attachment, the same asset arriving from three intermediaries with different numbers attached. Very little of it is structured. A good share of it is duplicated. Before anyone can form a view on a deal, somebody has to turn all of that into something comparable. ## What is an acquisition pipeline, in practice? In theory it is the set of opportunities under evaluation and their stage. In practice it is whatever state your inbox is in, plus a spreadsheet that someone updates. That gap matters, because the work of getting from one to the other is where most acquisition capacity actually goes. A pipeline is only as good as the point at which deals enter it, and for most teams deals enter it by hand. ## Who pays the normalisation tax? Usually an analyst. Identifying the asset, reconciling which of the three versions is current, lifting figures out of an attachment and putting them into a model is work that requires care and almost no judgment. It also scales linearly with deal flow, which means the busier the market, the more of your analytical capacity goes into data entry. The cost is not only time. Duplication is where error enters a pipeline. The same asset assessed twice under slightly different assumptions, or a stale set of numbers carried forward because nobody noticed a newer version had arrived, are both failures of housekeeping rather than of analysis. Both change decisions. ## What does it mean to start the pipeline where the deal arrives? It means treating arrival as the first step of the pipeline rather than as the thing that happens before it. Three things happen automatically as each opportunity lands: - **Extraction.** The asset, unit count, asking terms and figures are pulled out of the inbound itself, whatever form it arrived in. - **Duplicate resolution.** The same asset from multiple intermediaries is recognised as one opportunity with several sources, not three competing records. - **Scoring against your own criteria.** Not a generic screener's view of a good deal, but the things your investment committee actually weights, applied consistently to every deal before anyone opens it. The last point is the one most often conceded. Every deal screener encodes somebody's thesis. If it is not yours, the pipeline is being made consistent about the wrong things. ## Who should chase the missing information? Most deals arrive incomplete. A rent roll without unit detail. A T-12 missing a month. No service history at all. What follows is a week of correspondence: an analyst emailing a broker for the basics, following up, and reconciling what comes back. It is necessary work and a poor use of an expensive person. Workflows can carry it instead. Requests go out against a checklist you define, responses come back in, and each deal fills itself in, with the gaps visible rather than discovered late. The analyst's attention moves to the part that cannot be automated, which is deciding which deals deserve it. ## Why don't the financials tell you whether income is durable? Because a trailing twelve-month statement confirms that income existed. It cannot tell you whether that income survives under new ownership, and the financials are silent on the difference. Two assets can present identically in a model: same NOI, same occupancy. One has stable resident sentiment, a maintenance backlog under control and renewals holding. The other has deteriorating sentiment, a backlog that has been growing for a year and renewals propped up by concessions. Same page in the model, materially different hold period. (We have written separately on [underwriting a multifamily deal beyond the financials](/resources/underwrite-multifamily-beyond-financials).) The evidence that separates them is operational and it is unstructured: work orders, service history, response times, the texture of how the asset has actually been run. Historically it was left out of diligence for a practical reason rather than a philosophical one. Reading a year of that material on a deal timetable was not feasible for anyone. That constraint is what has changed. A note on where the line sits. This is asset-level evidence about how a property has been operated. It is not resident-level assessment, it has no place in screening decisions about individual residents, and we do not build it for that. ## What happens to diligence work at closing? Usually it is abandoned. The model goes into the fund file. The operating detail stays in somebody's notes. The team taking the asset over starts again from the rent roll, rebuilding an understanding that already existed three weeks earlier. That is costly at precisely the wrong moment. Resident support requests double in the first month after an acquisition. More than 40% of inherited staff leave within ninety days. Properties brought onto the owner's own systems inside that window contribute to fund returns faster than those left as they were found. We set out that window in more detail in [ensuring success of a property acquisition within the first 90 days](/resources/ensuring-success-of-a-property-acquisition-within-the-first-90-days). The fix is continuity rather than a handover document. If the context assembled during diligence already sits on the platform the team will use to run the asset, nothing has to be transferred at all. The people taking over inherit what was learned, including the specific problems diligence identified. ## How Travtus fits Every step described here can be bought as a point solution. Deal-flow trackers exist. Document extraction exists. None of it connects, which means each addition is another integration, another set of credentials, and another place where your thesis is encoded by somebody else. That is the [platform versus point solution](/resources/platform-vs-point-solutions) question, and in acquisitions it shows up as continuity: a deal tool stops at closing, because closing is where its job ends. Travtus is the [Everyday(AI)™](/resources/everyday-ai-property-management) platform for multifamily. Your own teams describe how a deal should arrive prepared, which signals to weight, what context to attach and what should follow it into the hands of the people who will run the asset, and they build it themselves, without an engineering programme and without replacing the systems already in place. Yours to build, in a sentence. ## Frequently asked questions ### Can deal information be extracted from broker emails and attachments automatically? Yes. Extraction runs on the inbound itself, so the asset, terms and figures are captured whatever form they arrived in. The same asset arriving from several intermediaries is resolved into one opportunity with multiple sources rather than several competing records. ### How do you score deals against your own investment criteria rather than a vendor's? By building the scoring logic yourself. A generic screener applies the criteria its vendor chose, which is why it produces consistency about the wrong things. On a platform, the people who own the thesis describe what should be weighted and the pipeline applies it to every deal before anyone opens it. ### How is operational evidence used in diligence without straying into resident screening? The evidence is asset-level: work orders, service history, response times and how the property has been run. It informs a view on whether income is durable. It is not used to assess individual residents and it plays no part in screening decisions. ### Does the diligence work carry over after close? That is the argument for doing it on a platform rather than in a deal tool. The context assembled during diligence stays available to the team that takes the asset on, so the first ninety days start from what was already learned rather than from the rent roll. **Want to see what a deal pipeline built to your own criteria looks like?** [Explore the Travtus platform](/resources/platform-vs-point-solutions) and how it connects deal-side context to the teams who run the asset after close. ## AI Agents for Property Management: What They Are and Where They Help URL: https://www.travtus.com/resources/ai-agents-for-property-management Date: 2026-07-15 Summary: A clear, vendor-neutral definition of AI agents for property management — what they actually are, where they help most, and how they differ from a scatter of point tools. > **An AI agent for property management is software that understands a request, decides what to do, and takes action across your systems on its own — reading a resident message, classifying it, and opening the right work order, for example. Agents help most on high-volume, repetitive work, and they deliver the most when built on one governed platform rather than bought as standalone tools.** The phrase "AI agent" gets used to mean everything from a chatbot to a science-fiction robot, which makes it nearly useless in a buying conversation. So let me be precise, because the definition is where most bad decisions start. I have watched operators buy an "agent" that turned out to be a canned FAQ, and others dismiss agents entirely after one narrow tool underdelivered. Both mistakes come from a fuzzy definition. Here is the plain version, and where agents actually earn their place in a property-management operation. ## What is an AI agent for property management? An AI agent for property management is software that understands a request, decides what to do about it, and takes action across your systems — without a person clicking through each step. That three-part loop is the whole distinction. A dashboard shows you something. A chatbot answers a question. An agent understands, decides, and *acts*. Concretely: a resident sends a message at 11pm saying the kitchen sink is leaking. An agent reads the message, recognizes it as an urgent maintenance issue, checks the unit and lease context, opens a correctly categorized and prioritized work order, and confirms back to the resident — before anyone on-site has opened their laptop. No one triaged it by hand. The reason this matters for buyers is that "agent" is a capability level, not a product category. What you should care about is how much of that understand-decide-act loop a given tool actually closes, and how safely it does so on your data. ## Where do AI agents help most in property management? Agents help most exactly where your teams are drowning: high-volume, repetitive, rules-based work. The clearest wins cluster in a few areas. - **Resident communication triage.** Reading inbound messages across email, text, and chat, understanding intent, and routing or answering. This is the highest-volume surface in most operations and the one that buries on-site staff first. - **Maintenance routing.** Turning a free-text complaint into a categorized, prioritized, correctly assigned work order — and summarizing the history when a tech picks it up. - **Conversation summarization and risk flagging.** Reading long resident threads, surfacing what matters, and flagging sentiment or compliance risk before it becomes a bad review or a complaint. - **Drafting and follow-up.** Producing consistent, on-brand first-draft responses that a person approves, so quality does not vary by who happens to be covering the desk. The pattern across all of these is the same: the agent handles the volume underneath, and the person handles judgment, exceptions, and relationships on top. Operators building these workflows on a platform report roughly 95% automation on the specific workflows they build and retention gains in the range of 25%, because service quality scales without adding headcount. This is a close cousin of the broader move toward [agentic AI for multifamily operators](/resources/agentic-ai-multifamily-operators) — worth reading if you want the operator's-eye view of where this goes. Agents do *not* help where the work is genuinely novel, relationship-heavy, or requires human accountability. Naming that boundary honestly is part of buying well. ## Should you buy standalone AI agents or build them on a platform? Here is the enterprise buying question I care about most: do you assemble a shelf of standalone agents, or build them on one platform? A standalone agent automates a single task, holds its own slice of data, and — this is the part that bites later — rarely changes how the wider company operates. Buy five of them and you have five contracts, five logins, five partial views of the same resident, and no way to govern the whole thing. This is the point-tool trap, and it is why so many agent pilots stall before they reach the portfolio. I have written before about [platform versus point solutions](/resources/platform-vs-point-solutions); agents are simply the sharpest current example of it. A platform inverts the economics. You build agents and workflows on shared, governed data that is already connected to your property management system, so each new build reuses the last one instead of starting over. The logic sits with your business users, not in a vendor's black box, and the IP you create stays yours. That is the difference between owning eleven experiments and owning a capability. For the fuller framing, [everyday AI for property management](/resources/everyday-ai-property-management) lays out what that operating posture looks like day to day. ## How Travtus approaches this Travtus treats agents as something you build, not something you buy off a shelf. On the [Everyday AI™ Platform](/everydayai), the loop runs **Connect → Explore → Create → Act.** Connect reads the systems you already run — Yardi, RealPage, Entrata, AppFolio — with no rip-and-replace. [Explore](/explore) lets your teams interrogate that connected data in plain language. Create is where you build the primitives that agents run on — Reports, Profiles, Scores, and Workflows — by describing what you want rather than writing code. Act is where those builds take action across the portfolio. Because building means describing, the people who understand leasing, maintenance, and service become the ones who build the agents — and enterprise-grade governance is carried in every feature, so it scales safely. The whole thing ladders to one idea: become an AI-native housing enterprise on one platform. Your agents are yours to build, on your data, by your people. Start with the [platform overview](/platform) or explore the [Everyday AI & the AI-Native Operator hub](/resources/category/everyday-ai). ## Frequently asked questions **What is an AI agent for property management?** Software that understands a request, decides what to do, and takes action across your systems — for example reading a resident message, classifying it, and opening the right work order — instead of waiting for a person to do each step. **Where do they help most?** High-volume, repetitive, rules-based work: triaging resident messages, routing maintenance, summarizing conversations, and drafting responses. Operators on a platform see roughly 95% automation on the workflows they build. **Are they the same as chatbots?** No. A chatbot talks; an agent acts across your connected systems. The value is in the doing, not the talking. **Do they replace staff?** No. They remove repetitive work so teams handle more without adding headcount, lifting retention by around 25% by scaling service quality. **Standalone agents or a platform?** A platform, for any enterprise — agents built on shared governed data compound, while standalone tools stall. *See how agents get built on the [Everyday AI™ Platform](/platform), or [book a demo](/demo) to see it on your data.* ## AI Resident Comms That Don't Feel Like a Bot URL: https://www.travtus.com/resources/ai-resident-comms-that-dont-feel-like-a-bot Date: 2026-07-15 Summary: Residents don't leave because you use a property management chatbot AI assistant. They leave when they feel ignored. Tone and the right human handoff are everything. > **A property management chatbot AI assistant only earns its keep when it feels human: warm in tone, instant on the routine questions, and honest about its limits. The moment a request needs judgment, it should hand off to a real person with the full story already attached, so the resident never has to start over.** I spent years running centralized resident-services and leasing teams, and I can tell you the page that came in at 9 p.m. was almost never "I'm leaving because of the rent." It was "I've messaged three times about my parking spot and nobody's gotten back to me." Residents don't churn over a bot. They churn over feeling ignored. So when operators ask me whether they should worry that automation will make their communities feel cold, I flip the question. The cold experience is already here: it's the voicemail that goes unreturned, the leasing email answered on day four, the maintenance follow-up that never comes. A good AI assistant is how you make service feel *warmer*, not colder, because it closes the gaps where residents currently feel dropped. The trick is building one that sounds like your community and knows exactly when to step aside for a person. ## What makes an AI assistant feel like a bot instead of a person? It feels like a bot when it does three things: it answers a question nobody asked, it loops the resident in circles, and it refuses to let them reach a human. Every "I'm sorry, I didn't understand that" is a small betrayal of trust. What makes it feel human is the opposite. It reads the actual message. It answers the real question in plain, warm language. And when it can't help, it says so and connects the resident to someone who can, without making them repeat a word. Picture a resident who texts at 11 p.m.: "My AC died and it's 88 degrees in here." A bot-feeling system replies with a link to the resident portal and a list of office hours. A human-feeling assistant says, "That sounds miserable, I'm sorry. I've logged an emergency work order and flagged it as urgent, someone will be in touch tonight," and then it actually does both. Same automation. Completely different relationship. The renter-facing Customer Service Agent we build answers across chat, SMS, email, and voice precisely so that message gets a real answer at 11 p.m., not at 9 a.m. the next morning. ## How does AI know when to hand off to a human agent? This is the question that decides whether residents trust the whole system, so let me answer it directly. A well-designed assistant hands off on three signals: emotion, complexity, and risk. If a resident is upset, if the request is unusual enough that a wrong answer would do harm, or if the reply carries financial or legal weight, the assistant should stop and route to a person. The part most tools get wrong is *what* they hand over. A warm transfer is not "please hold while I connect you." It's the human teammate receiving the entire conversation, the resident's history, and the context of what's already been tried, before they say hello. Our Customer Service Agent hands off to a human *with full context* for exactly this reason. Nothing frustrates a resident faster than explaining the whole story twice. There's a layer before that, too. A Guardrail Agent screens input before the agent acts, so if something looks sensitive, off-policy, or likely to produce a bad answer, it never gets that far, it's caught and escalated first. Think of it as the difference between a new hire who checks with a manager before promising a fee waiver, and one who just says yes. Residents feel the difference even when they can't name it. I've watched leasing teams relax once they trust the handoff. When they know the assistant will always pull them in for the conversations that matter, and always arrive with context in hand, they stop babysitting the inbox and start doing the work that actually renews leases. ## How do you keep the tone warm and consistent across every channel? Tone is not a nice-to-have in resident comms, it *is* the product. A resident who gets a fast, kind answer at midnight remembers your community as the place that had their back. A resident who gets a curt, robotic reply remembers you as the place that couldn't be bothered. Consistency is the harder half. Your best leasing agent is warm on a Tuesday morning and warm again on a Friday night. Most teams can't hold that line at 2 a.m. or across a thousand messages a week, not because they don't care, but because they're human and stretched thin. This is where an AI assistant genuinely helps: it delivers the same considered, on-brand voice on message one and message ten thousand, on chat and on voice, at noon and at midnight. The way you get there is deliberate. You give the assistant your community's real information so it answers accurately. You set the tone on purpose, the words, the warmth, the level of formality your residents expect. And you watch. We track Conversation Scores and Talent Scores so you can see the quality of what's actually being said, spot where the voice is drifting, and correct it, the same way a good service leader coaches a team. Fair, neutral language for every resident isn't just good manners; it's how you keep communications consistent and defensible. ## How Travtus approaches this We build resident communications on the [Everyday AI™ Platform](/everydayai) so the assistant, the guardrails, and the resident record all work from the same source of truth, not a stack of disconnected point tools. The renter-facing Customer Service Agent lives inside our [Resident Experience](/solutions/resident-experience) solution: it answers residents and prospects across chat, SMS, email, and voice, and hands off to your team with the full conversation and history attached. Underneath, [Response Management](/response-management) is what makes the handoff clean, so a human teammate always picks up with context instead of a cold "how can I help you?" The Guardrail Agent screens input before the agent responds, and the Customer Profile gives both the AI and your people the full story on a resident before anyone replies. Operators like Cortland, MAA, BH, Continental, and Preiss use this approach, and it's part of how communities have seen retention lift by roughly 25%. None of this is a rip-out-and-replace project. The tone, the escalation rules, and the voice residents hear are yours to build, shaped around how your communities actually talk. For the wider picture, the [Resident Experience & Communications hub](/resources/category/resident-experience) pulls the rest together, and it pairs naturally with our take on [conversational AI as a game-changer for multifamily operations](/resources/harnessing-the-power-of-conversational-ai-a-game-changer-for-multifamily-operations) and on [catching resident complaints before they become bad reviews](/resources/catch-resident-complaints-before-bad-reviews). ## Frequently asked questions **What is a property management chatbot AI assistant?** It's an AI agent that answers residents and prospects across chat, SMS, email, and voice, day or night. The good ones resolve routine requests instantly and, when a question needs judgment, hand off to a human teammate with the full conversation and resident history attached, so nobody has to repeat themselves. **How does AI know when to hand off to a human agent?** A well-built assistant recognizes emotion, complexity, and risk. When a resident is upset, the request is unusual, or a promise carries financial or legal weight, it stops and routes to a person. At Travtus, a Guardrail Agent screens input before the agent acts, so the handoff happens before a bad answer ever reaches the resident. **Do residents mind talking to an AI assistant?** Residents mind being ignored, not being helped. Our customer results show people are happy with an AI assistant when it's fast, accurate, and warm, and when a human is one step away for anything it can't solve. The frustration comes from dead ends and canned replies, not from automation itself. **Will an AI assistant replace my leasing and resident services team?** No. It removes the repetitive volume, the after-hours messages and the same twenty questions, so your team spends its time on the conversations that actually build renewals. The assistant carries the routine load; your people carry the relationships. **How do you keep an AI assistant on-brand and fair?** Give it your community's real information, set the tone deliberately, and screen every input before the agent responds. Keep language neutral and consistent for everyone, escalate anything sensitive to a person, and review conversation quality regularly so the voice residents hear always sounds like your community. *See how warm, always-on resident comms come together in [Resident Experience](/solutions/resident-experience), then [book a demo](/demo).* ## Stop Drowning the Data Team in Ad-Hoc Report Requests URL: https://www.travtus.com/resources/automated-report-generation-real-estate Date: 2026-07-15 Summary: When every report is a ticket, analysts become a lookup service and never get to the real work. Here is how automated report generation in real estate scales analytics without more headcount. > **Automated report generation in real estate scales analytics without more headcount by removing three kinds of work from the data team: routine lookups move to governed self-service, recurring reports become scheduled and self-delivering, and downstream systems get data feeds. What remains is the novel analysis only analysts can do — so capacity grows by subtracting work, not adding people.** I've been the analyst buried under the request backlog, and I've managed the team that was. The pattern is identical everywhere. A skilled analyst — someone who could be modeling churn or pressure-testing a strategy — spends the majority of the week pulling the same handful of numbers for people who can't pull them themselves. Occupancy by region. Delinquency by community. The owner package, again. It's a waste of the most expensive and scarce resource in the building, and it never ends, because demand for answers grows faster than any team you can hire. The instinct is to fix it with headcount: more requests, more analysts. But that just scales the bottleneck. If each analyst's week is mostly repetitive lookups, adding a second analyst buys you more lookups, not more insight. The requests keep pace, and the deep work — the analysis that actually moves decisions — stays perpetually at the bottom of the list. You can't hire your way out of a work-design problem. ## How to reduce ad-hoc reporting requests Separate the request pile into three types, because each has a different fix and lumping them together is why the backlog feels unbeatable. - **Routine one-off lookups** — "what's occupancy at these five communities?" These don't need an analyst at all; they need to be answerable by the person asking. Governed self-service handles them. - **Recurring reports** — the weekly health summary, the monthly owner package. These are the same report on a timer. They should generate and deliver themselves, not be reassembled by hand every cycle. - **Downstream data needs** — another system that needs your data on a schedule. That's a data feed, not a report request, and it shouldn't hit a human at all. Here's the arithmetic that makes the case. Say the data team fields 100 requests a week. In most operations, something like 60 are routine lookups, 30 are recurring reports, and only 10 are genuinely novel analysis. Route the 60 to self-service and automate the 30, and the team's inbound drops from 100 to 10 — and those 10 are the interesting ones. You haven't changed headcount. You've changed what the headcount does. That's how you reduce ad-hoc requests: you stop treating three different problems as one queue of tickets. ## How to scale analytics without hiring more analysts You scale by raising each analyst's coverage, and you raise coverage by automating the repetitive layer. Automated report generation is the core of it: you describe a report once — what it contains, who it's for, how often — and it generates and delivers on schedule without anyone touching it. The weekly portfolio-health report lands Monday morning. The monthly owner package assembles itself. The daily exception report flags the outliers before anyone asks. Two things make this practical rather than aspirational. First, the reports are built by describing them, not by engineering them — a business user or analyst can stand one up in plain language, so creating a new scheduled report isn't itself a project. Second, once the recurring load is automated, the same self-service layer lets business users answer their own one-off questions in plain language, with cited answers, so those never become tickets either. The teams working this way see roughly a 15% productivity lift — and the more valuable effect is qualitative: analysts get their week back. A worked contrast. Before: an analyst spends Monday rebuilding the portfolio summary, Tuesday and Wednesday on lookups, and gets to real analysis Thursday if nothing breaks. After: the summary delivered itself, the lookups were self-served, and the analyst spends the week on the churn model leadership has been asking for since spring. Same person, same salary, completely different output. That's scaling analytics without scaling the org chart. ## Is self-service reporting safe, or does data quality suffer The honest answer is that ungoverned self-service is dangerous, and that's the objection worth taking seriously. Hand everyone a query tool with no shared definitions and you get five versions of "occupancy," each computed slightly differently, argued over in the next meeting. That's worse than a backlog — it's a trust problem. The fix isn't to lock the data back up; it's governance under the self-service. When metrics are defined once on a shared layer, and every answer carries citations back to its source, opening up access doesn't fragment the truth — everyone pulls from the same definitions and can see where a number came from. Governance is precisely what lets you widen access safely. Citations do double duty here: they let a business user trust the answer, and they let the data team audit what's being asked and answered without sitting in the middle of every request. Done this way, self-service raises data quality, because the definitions are centralized instead of re-invented in a hundred private spreadsheets. ## How Travtus approaches this Travtus removes the reporting burden as part of the [Everyday AI™ Platform](/everydayai). You build [Reports](/reports) by describing them and schedule them to generate and deliver on their own; Data Feeds push data to downstream systems without a ticket; and Explore lets business users answer their own questions in plain language, with citations, reasoning across conversations, records, and everything published. You also build the other primitives — Profiles, [Scores](/scores), Workflows — the same way, by describing them, so standing up new reporting isn't an engineering project. It runs on your existing systems with no rip-and-replace, which is why operators like Cortland, MAA, and Preiss could shift analyst time toward high-value work rather than lookups. For the broader picture see [Customer Intelligence](/customer-intelligence), the [Data, Analytics & Portfolio Intelligence hub](/resources/category/data-analytics), and how AI [boosts apartment operations and financial performance](/resources/using-ai-to-boost-apartment-operations-and-financial-performance). ## Frequently asked questions **How do I reduce ad-hoc reporting requests?** Split the pile: route routine lookups to governed self-service, convert recurring requests into scheduled reports, and push downstream needs to data feeds. What's left is genuinely novel analysis — a fraction of the original volume. **How do I scale analytics without hiring more analysts?** Stop scaling headcount with request volume. Automate the repetitive work so each analyst covers far more surface area — you add capacity by removing work, not adding people. **What is automated report generation in real estate?** Producing recurring reports without a person assembling them each time. You describe the report once, set a schedule, and it generates and delivers on its own. **What work should analysts do instead?** The work only they can do — data modeling, metric governance, and the deep analysis behind real decisions. Ad-hoc lookups are the lowest-value use of a skilled analyst. **Is self-service reporting safe?** It's safe when it's governed. Define metrics once on a shared layer and carry citations on every answer, and opening up access strengthens the single source of truth rather than fragmenting it. *Free your analysts for the work that matters with automated [Reports](/reports) — [book a demo](/demo).* ## Catch Resident Complaints Before They Become Bad Reviews URL: https://www.travtus.com/resources/catch-resident-complaints-before-bad-reviews Date: 2026-07-15 Summary: The gap between a private complaint and a public one-star review is where retention is won. Online reputation management for apartment communities starts before the review. > **The best online reputation management for apartment communities happens before a review is ever posted. Track resident sentiment daily as an always-on Score, surface the red flags early, and resolve the complaint while the resident is still yours. A public one-star review is almost always a private complaint that nobody caught in time.** Here's a pattern I saw over and over running resident-services teams: the one-star review that blindsided the community manager was never actually a surprise. The warning signs were all there, a testy maintenance exchange, a renewal question that went unanswered, a "this is the third time I've asked" text. We just didn't have a way to see them until the resident said it out loud, in public, where every future prospect could read it. That gap, between the private frustration and the public review, is the most valuable and most neglected window in resident experience. Close it, and you don't just protect your rating; you keep a resident. Miss it, and you spend the next six months apologizing on Google while your leasing team works twice as hard to replace someone you already had. ## How do you catch resident complaints before they become bad reviews? You stop relying on the moment a resident decides to broadcast their frustration, and you start listening long before that. Practically, that means treating sentiment as something you measure every day, not something you discover at renewal or read in a star rating. At Travtus we do this with Scores: an always-on satisfaction and sentiment pulse, scored 0 to 100 daily, for every resident. I think of it as a new-age NPS. Instead of one blunt number pulled from a survey a resident may never open, you get a living read on how each resident actually feels, updated as they interact with your community. When a Score dips, that's your red flag, and it lands in your system days or weeks before it would ever land on Google. Consider a resident whose sentiment slides from the low 80s to the high 40s over two weeks after a slow maintenance repair. On the old model, you'd find out at renewal when they gave notice, or online when they warned everyone else. With a daily Score, that resident surfaces now, while a phone call and a real fix can still change the ending. ## What signals tell you a resident is about to leave a bad review? The strongest signals are almost always operational, not financial. Repeated contact about the same unresolved issue. A sharp change in tone. A maintenance request that's been reopened. A prospect-turned-resident whose early enthusiasm has gone quiet. Individually, any of these is easy to miss on a busy Tuesday. Together, they're a resident writing a review in their head. The problem was never that operators don't care, it's that these signals are scattered across a portal, an inbox, a work-order system, and a dozen conversations no one person ever sees end to end. So we pull them into one place. The Customer Profile gives your team the full story on a resident, every conversation, every issue, every follow-up, so the pattern is visible instead of buried. The Newsfeed surfaces feedback and reviews as they come in, so nothing sits unseen. And when feedback points to a recurring problem, Action Plans auto-generate from that feedback, turning "something's wrong" into a concrete list of what to do about it. I've watched a community manager use exactly this to catch a cluster of complaints about package handling before a single one hit a review site, fix the process, follow up with the residents who'd flagged it, and turn a brewing reputation problem into a set of renewals. That's the whole game. ## What is online reputation management for apartment communities, really? Most people hear "reputation management" and picture responding to reviews after they post, drafting the polite reply to the angry one-star, nudging happy residents to leave a five. That work matters, and you should do it well. But it's the last ten percent. The real work is upstream. Genuine online reputation management for apartment communities is a loop: listen to sentiment continuously, resolve problems while they're still private, follow up so the resident feels heard, and *then* invite the now-happy resident to share their experience publicly. Do that and your public rating stops being something you defend and starts being an honest reflection of communities that actually take care of people. The reviews take care of themselves when the residents are taken care of first. And when a review does post, speed and sincerity decide how it reads to the next prospect. A fast, specific, human response, not boilerplate, tells a hundred future renters that this is a community that listens. The Newsfeed makes sure the review reaches your team quickly enough for that response to feel genuine rather than late. ## How Travtus approaches this We build all of this on the [Everyday AI™ Platform](/everydayai), so resident sentiment, conversations, and feedback live together instead of in separate tools that never talk. Scores give you the daily 0-to-100 pulse; the Customer Profile holds the full story on each resident; the Newsfeed surfaces feedback and reviews as they land; and Action Plans turn that feedback into concrete steps, so a red flag becomes a fix instead of a note nobody reads. This sits at the heart of our [Resident Experience](/solutions/resident-experience) solution, and it's tied closely to [Response Management](/response-management), because catching a complaint only matters if someone responds well and fast. Operators like Cortland, MAA, BH, Continental, and Preiss work this way, and it's part of how communities have seen retention lift by roughly 25%, you keep more of the residents you already fought to win. The rules for what counts as a red flag, and how your team acts on it, are yours to build around how your communities actually run, which makes the whole thing feel like an extension of your team rather than a bolt-on. For more, the [Resident Experience & Communications hub](/resources/category/resident-experience) gathers the rest, and this pairs well with [AI resident comms that don't feel like a bot](/resources/ai-resident-comms-that-dont-feel-like-a-bot) and with turning [customer support into customer success for multifamily operators](/resources/transforming-customer-support-to-customer-success-for-multifamily-operators). ## Frequently asked questions **How do you catch resident complaints before they become bad reviews?** Stop waiting for the annual survey or the star rating. Track resident sentiment every day as an always-on signal, so a frustrated resident shows up as a red flag in your system before they show up on Google. Then act while they're still your resident, resolve the issue, follow up, and turn a would-be one-star into a renewal. **What is online reputation management for apartment communities?** It's the practice of protecting and improving how your community is seen publicly, but the real work happens privately, before anyone posts. It means listening to resident sentiment continuously, resolving problems early, encouraging happy residents to speak up, and responding to reviews with genuine care rather than boilerplate. **Why isn't an annual resident survey enough?** A once-a-year survey is a lagging indicator. By the time you read the results, the frustrated resident has already decided to leave or already posted the review. An always-on sentiment Score works like a new-age NPS, giving you a daily pulse so you can intervene in real time instead of reading about the problem months later. **Can you really change a resident's mind before they leave a bad review?** Yes, and it happens all the time. Most residents post a public review only after they feel a private complaint was ignored. Catch the frustration early, respond with a real fix and a genuine follow-up, and the same resident who was about to warn the internet often renews instead. Speed and sincerity are everything. **How does AI help with resident sentiment and reviews?** AI reads every conversation, work order, and message as it happens, scores sentiment daily, and surfaces the residents who are slipping before they churn. It can auto-generate an action plan from the feedback and route reviews to your team fast, so nothing sits unanswered. Your people still own the relationship, the AI just makes sure nothing falls through. *See how daily sentiment and fast response protect your communities in [Resident Experience](/solutions/resident-experience), then [book a demo](/demo).* ## How to Make a Multifamily Company AI-Native (Without Ripping Out Your Stack) URL: https://www.travtus.com/resources/how-to-make-a-multifamily-company-ai-native Date: 2026-07-15 Summary: A practical playbook for becoming an AI-native multifamily operator — the four goals of an AI-native enterprise, why no rip-and-replace is required, and how business users own the logic. > **You make a multifamily company AI-native by adopting one platform, connecting it to the property management systems you already run, and letting business users build their own reports, profiles, scores, and workflows by describing what they want. There is no rip-and-replace — you keep Yardi, RealPage, Entrata, or AppFolio and add a governed AI operating layer on top. AI-native is an operating posture, and it is yours to build.** Every COO I meet asks the same question within the first ten minutes: "Do I have to change everything?" It is the right question, and the honest answer is no. The operators who went AI-native in the last year did not rip out their stack, retrain everyone at once, or hand control to a vendor. They did something more disciplined and less dramatic. Being AI-native is not a feature you buy or a project you finish. It is how the company operates — everyday work understood, surfaced, and automated by default. Here is the playbook for getting there without tearing anything out. ## What are the goals of an AI-native multifamily enterprise? There are four goals, and they map cleanly onto how the work actually flows. Treat them as the destination, not a checklist to rush. 1. **Connect the data.** Get to one governed picture of your residents, properties, and conversations — drawn from the systems you already run, not copied into yet another silo. Everything downstream depends on this. 2. **Explore in plain language.** Let any authorized team member ask questions of that connected data in plain English and get an answer, instead of filing a report request and waiting a week. This is where the productivity shows up first — operators see gains around 15% here as manual digging disappears. 3. **Let business users build the logic.** The people who understand leasing, maintenance, and service should be the ones building the automation — by describing what they want, not by writing code. This is the goal most operators underestimate and the one that separates AI-native companies from AI-curious ones. 4. **Act across the portfolio.** Put those builds to work everywhere at once, with governance carried into every action, so a good idea from one region becomes standard practice everywhere. Hit all four and the character of the company changes: work that used to bury on-site teams gets handled underneath them. Operators building this way report roughly 95% automation on the workflows they create and retention gains in the range of 25% as service quality scales without new headcount. Those numbers are the reward for pursuing the goals in order — not for buying more tools. ## Which AI platform integrates with my PMS without rip-and-replace? This is the fear that stalls most AI-native efforts before they start, so let me address it directly: you do not replace your system of record. A platform is compatible when it *reads from* your existing systems rather than demanding you migrate off them. Practically, that means the platform connects to Yardi, RealPage, Entrata, and AppFolio and operates on top of them. Your PMS stays exactly where it is and keeps doing what it does. The AI operating layer sits above it, drawing on that connected data. Nothing gets torn out, no multi-year migration, no frozen roadmap while a replacement is stood up. This is not a minor convenience — it is the whole reason AI-native adoption can be incremental. You connect one data source, build one workflow, prove the value, and expand from there. Late movers actually benefit here: the [clean-slate advantage](/resources/the-clean-slate-advantage-why-late-adopters-may-win-the-ai-race) means you can skip the dozen-tool sprawl entirely and go straight to a platform. And because there is no rip-and-replace, the risk of starting is low and the cost of expanding compounds in your favor. ## Do business users or IT build the AI logic? Business users build the logic. This is the shift that makes a company genuinely AI-native rather than a place that owns AI tools, and it is worth being emphatic about. For most of enterprise software history, the people who understood the work and the people who could change the software were two different groups, separated by a ticket. An operations leader who knew exactly how a renewal conversation should be handled had to describe it to IT, wait, and hope the result matched. On a modern platform that gap closes, because building means *describing what you want in plain language* — no code. The regional manager builds the workflow. The resident-services lead builds the score. The person closest to the work owns the logic. IT does not disappear; it changes role. It sets the governance guardrails — who can see what, what is auditable, how data is handled — and then gets out of the way of building. That division is the practical meaning of enterprise-grade governance carried in every feature. It also settles the ownership question that matters most to a CEO: the logic your people build is your intellectual property, and it stays yours. You are not renting someone else's black box; you are compounding an asset. For the broader case on why this beats assembling point tools, [platform versus point solutions](/resources/platform-vs-point-solutions) is the argument in full. ## How Travtus approaches this Travtus built the [Everyday AI™ Platform](/everydayai) to make this playbook runnable without a rip-and-replace. The loop is **Connect → Explore → Create → Act**, and it maps one-to-one onto the four goals above. Connect reads Yardi, RealPage, Entrata, and AppFolio in place. [Explore](/explore) lets your teams question that connected data in plain language. Create is where business users build the primitives — Reports, Profiles, Scores, and Workflows — by describing what they want, not by coding. Act puts those builds to work across the portfolio, with enterprise-grade governance in every feature. That is what "AI-native" means in practice, and it is why the message is simply: become an AI-native housing enterprise on one platform. The result is yours to build — on your data, by your people, with your IP staying yours. Start with the [platform overview](/platform), or work through the [Everyday AI & the AI-Native Operator hub](/resources/category/everyday-ai) for the rest of the playbook. ## Frequently asked questions **How do you make a multifamily company AI-native?** Adopt one platform, connect it to the systems you already run, and let business users build their own reports, profiles, scores, and workflows on that shared data. It is an operating posture, not a feature switch. **Which platform integrates with my PMS without rip-and-replace?** One that reads from your existing systems. The Everyday AI Platform connects to Yardi, RealPage, Entrata, and AppFolio and operates on top of them — you keep your system of record. **What are the four goals?** Connect the data, explore it in plain language, let business users build the logic by describing it, and act across the portfolio. **Who builds the AI logic?** Business users — they describe what they want in plain language while IT sets the governance guardrails. **How long does it take?** It is incremental — start with one connected source and one workflow, prove value, and compound from there. *See the playbook run on the [Everyday AI™ Platform](/platform), or [book a demo](/demo) to map it to your stack.* ## Property Management Automation: What to Automate First URL: https://www.travtus.com/resources/property-management-automation-what-to-automate-first Date: 2026-07-15 Summary: Start property management automation with the high-volume, low-judgment work that buries your teams — intake, routing, reminders, status — and build it by describing it, not coding it. > **Automate the high-volume, low-judgment work first: request intake, routing, reminders, and status updates. This work repeats constantly, follows clear rules, and buries your teams without needing their judgment — so it returns hours fast and carries almost no risk. Build each workflow by describing it in plain language, on your own data, and keep it yours to change.** It's Tuesday. Two techs called out, the regional wants a portfolio walk by Thursday, and the shared inbox has 40-odd messages that nobody has sorted yet. Half are duplicate work-order requests, a handful are genuinely urgent, and three are residents who emailed on Friday and still haven't heard back. Your community manager is going to spend the first two hours of the day just figuring out what's in the pile before a single problem gets solved. That is the real starting point for property management automation. Not a dashboard, not a strategy deck — the pile. And the good news is that most of what's in it is exactly the kind of work a machine should have handled before anyone sat down. You don't have to automate everything to feel relief. You have to automate the right things first. ## What should you automate first in property management? Start with work that is high-volume and low-judgment. That combination is the whole trick. High-volume means it happens constantly, so even a small time savings per instance adds up to real hours. Low-judgment means the rules are clear enough that you'd be comfortable letting a well-trained new hire handle it — which means you'd be comfortable letting software handle it too. Four categories fit that description in almost every portfolio: - **Intake.** Turning inbound requests — emails, calls, portal messages, texts — into structured, categorized items. No more re-reading the same message three times to figure out what it is. - **Routing.** Getting each item to the right person or team the first time. Maintenance to maintenance, a lease question to leasing, an escalation to the regional — automatically, based on what the request actually says. - **Reminders and follow-ups.** The nudges that fall through the cracks when a property is short-staffed: the renewal follow-up, the "did this get resolved" check, the vendor who said they'd call back. - **Status updates.** Telling the resident their request was received, is in progress, is done — without a human writing the same three sentences forty times a day. Skip, for now, the work that genuinely needs a person: the judgment call on a difficult resident, the exception nobody planned for, the decision that carries real money or real risk. Automating those first is how automation projects earn a bad name. Earn trust on the boring, repetitive volume, then expand. ## Why high-volume, low-judgment work is the safest place to start There's a reason this order matters beyond just time savings. When you automate low-judgment work, the cost of getting it slightly wrong is low. If an intake rule miscategorizes a message in week one, you notice, you adjust, and no one gets hurt. That's a very different bet than handing a machine a decision that affects a lease or a resident relationship on day one. It's also where the hours actually are. When operators map where their on-site teams' time goes, the surprise is never the hard, interesting problems — those are rare. It's the sheer volume of small, repetitive handling: sorting, forwarding, acknowledging, chasing. That's the water your teams are swimming in. Drain it, and you get people back for the work that needs a human. Travtus operators typically see around a 15% productivity gain from this alone, and it shows up first as teams who aren't drowning by 10 a.m. And when the low-judgment work runs itself, response times get faster and more consistent — which is quietly one of the biggest drivers of resident retention. People don't leave because you sorted their email slowly; they leave because "ignored" adds up. Removing that friction protects NOI through the cost side and the retention side at once, without touching a single thing about how you price or lease. ## How do you build these workflows without a developer? Here's the part that changed my mind about automation. For years, "automate that" meant a request to IT, a spot in someone's backlog, and a six-week wait for a workflow that didn't quite match how the work actually happens on-site. So nobody bothered, and the pile stayed. On a modern platform, you build a workflow by *describing* it. You say, in plain language: when a maintenance request comes in, categorize it, route it to the on-site team, acknowledge the resident, and flag anything marked urgent to the community manager. The platform builds it. No code, no ticket, no waiting on a developer who's never covered a property short-staffed. That matters for a reason beyond speed. The person who knows the process is the person who builds it — and can change it the moment the process changes. On Travtus, the automation you build runs on your data, and the logic stays yours to adjust. It's not a black box some vendor owns. It's yours to build, and yours to fix at 4 p.m. on a Tuesday when the process shifts. That ownership is what keeps automation useful a year later instead of stale. ## What "best AI for multifamily operational efficiency" actually means When operators ask me for the best AI for multifamily operational efficiency, they're usually expecting a single-tool recommendation. But a point tool automates one task; it doesn't change how the company operates. You end up with a dozen disconnected tools, each handling a sliver, none of them talking to each other — and you've just moved the coordination problem, not solved it. The better answer is range: one platform that connects to the systems you already run, works across intake, routing, reminders, and status, and grows as you find the next thing worth automating. It should layer onto Yardi, RealPage, Entrata, or AppFolio — not replace them. Owner-operators like Cortland, MAA, BH, Continental, and Preiss don't rip out their system of record to get efficiency; they build on top of it. That's the difference between an efficiency project that stalls at a pilot and one that becomes how the whole portfolio runs. ## How Travtus approaches this Travtus is built to automate the exact pile that buries your teams — and to let operators build it themselves. The [Everyday AI™ Platform](/everydayai) follows a simple path: Connect to your existing systems, Explore your data, Create the primitives you need, and Act. You build primitives — Reports, Profiles, Scores, and [Workflows](/workflows) — by describing them in plain language, on your own data, with your IP staying yours. No rip-and-replace of Yardi, RealPage, Entrata, or AppFolio. Start with the high-volume, low-judgment work, prove the hours back, then expand across the portfolio — the practical route to becoming an AI-native housing enterprise on one platform. For the bigger picture on pulling scattered work into one place, see [Centralization](/centralization) and the [Operations & Efficiency hub](/resources/category/operations-efficiency), or the deeper guides on [how to hyper-automate your residential operations and centralize](/resources/how-to-hyper-automate-your-residential-operations-and-centralize) and [platform vs. point solutions](/resources/platform-vs-point-solutions). ## Frequently asked questions **What should you automate first in property management?** Start with high-volume, low-judgment work: request intake, routing to the right team, reminders and follow-ups, and status updates. This work repeats constantly, follows clear rules, and eats hours without needing human judgment — so automating it returns time fast and carries almost no risk if you get it slightly wrong at first. **What is the best AI for multifamily operational efficiency?** The best fit is a platform that works across your existing systems rather than a single-task point tool. Look for one that connects to your data, lets you build workflows by describing them in plain language, and layers onto Yardi, RealPage, Entrata, or AppFolio without a rip-and-replace. That range is what compounds into real efficiency. **Do you need to replace your property management system to automate?** No. Modern platforms sit on top of your system of record — Yardi, RealPage, Entrata, AppFolio — and read from it rather than replace it. You keep the system your teams already know and add automation across it. Rip-and-replace is slow, risky, and usually unnecessary to start returning hours. **How do you build a workflow without writing code?** You describe the outcome you want in plain language — the trigger, the steps, who gets notified, what gets logged — and the platform builds the workflow. No scripting, no developer backlog. An operator who knows the process can build and adjust it directly, which is why the work you automate stays yours to change. **How much of the work can automation actually handle?** For well-scoped, repetitive workflows, automation can handle the vast majority of the volume end to end, with people stepping in only on the exceptions. Operators on the Travtus platform see roughly 95% workflow automation on the flows they build, freeing teams to spend their judgment where it matters. *See what's worth automating first with [Workflows](/workflows) you build by describing them — then [book a demo](/demo).* ## How to Get Real-Time Visibility Across Your Whole Portfolio URL: https://www.travtus.com/resources/real-time-portfolio-visibility-multifamily Date: 2026-07-15 Summary: Quarter-end is too late to learn an asset is sliding. Here is how portfolio risk monitoring in real estate moves from static reports to daily, ranked asset-health signals you can act on. > **Portfolio risk monitoring in real estate works when every asset carries a daily health score built from leading operational signals, ranked worst-to-best, with a plain-language explanation attached. That replaces the quarter-end surprise with a same-day read on which communities are sliding and why, so you act while the outcome is still yours to change.** I spent years building the portfolio reports the C-suite read every quarter. The reports were accurate, well-formatted, and reliably late. By the time a delinquency trend or a renewal problem showed up in the numbers I published, the operational cause was two to three months old. The asset manager's job is to protect value, and I was handing leadership a rear-view mirror. The gap that hurts you isn't a gap in data. Most portfolios are drowning in it — property management systems, work-order platforms, resident messages, survey tools. The gap is between all that transactional data and a decision you can make today. Real-time visibility isn't more dashboards. It's a way to compress fifty properties into the handful that need you this week, and then explain why. ## How to get real-time visibility across all my properties Start by putting every asset on one comparable scale. A portfolio of fifty communities produces fifty different reports, each true in isolation and useless in aggregate — you can't rank a rent roll against a maintenance log. The fix is a single normalized measure of asset health that every property carries. Community Scores do exactly this: a 0-100 health score for each community, refreshed daily. Because the scale is identical across the portfolio, you can sort the whole book worst-to-best in one view. That one design choice changes the reading pattern completely. Instead of opening fifty files hunting for problems, you read a ranked list top-down and stop when the scores stop worrying you. Here is the arithmetic that matters. Suppose you oversee 50 assets and can give real attention to five in a given week. Without ranking, you're sampling — you look at the assets you happened to hear about, which are usually the ones already on fire. With a daily ranked score, the five worst rise to the top automatically. You've moved from reacting to the loudest property to triaging the actual bottom decile. Same five hours of attention, aimed at the assets where it changes the outcome. Daily matters as much as ranked. A score that refreshes monthly still hides three to four weeks of drift. A daily refresh means a community that starts sliding on the 3rd is on your radar by the 4th, not at month-end. ## How do asset managers track portfolio risk in real time By watching the signals that move first. The risks that hurt a portfolio aren't in the financial statements yet — they're upstream, in how residents feel and how the operation is running. Three leading indicators consistently move weeks ahead of revenue: - **Resident sentiment.** A rise in frustrated messages or negative review language at a community shows up long before it shows up in renewals. Sentiment is the earliest tremor. - **Maintenance backlog.** Aging, unresolved work orders are a direct read on operational strain and a reliable predictor of dissatisfaction. A backlog that's growing week-over-week is a leading indicator of the churn that will land in occupancy two quarters out. - **Renewal behavior.** Softening renewal intent is the last warning before it converts into vacancy and turn cost. None of these appear on a variance report until they've already become a financial problem. Track them as a group and you get a genuine early-warning system. This is what separates real-time risk monitoring from real-time reporting: you're not watching the outcome faster, you're watching the causes. A worked example. Two communities show identical trailing occupancy at 94%. On the financials they're twins. But community A has a maintenance backlog up 40% over three weeks and rising negative sentiment; community B is flat on both. A trailing report treats them as equal. A leading-signal score separates them cleanly — A is a problem developing, B is stable — and tells you where to spend the site visit. That's the difference between managing risk and documenting it after the fact. ## Can I see why a property's health changed, not just that it did A number that moves is a question, not an answer. The score tells you community A dropped 11 points this week; it doesn't tell you whether that's a staffing gap, a delayed capital project, or one viral bad review. Visibility that stops at the number just relocates the investigation — now you're pulling logs and emailing the regional instead of reading a report. Two layers close that loop. Community Profiles hold the standing context for each asset — the operational narrative, the history, the specifics that make a score legible. And Explore, inside Gateway, lets you ask the "why" directly. You can type, in plain language, "why did the health score at [community] drop this week?" and get an answer that reasons across everything: resident conversations, the underlying records and numbers, and your published Reports, Profiles, and Scores. It uses SQL over your structured data, vector search over unstructured text, and web search where relevant, and it answers with citations so you can see exactly what drove the conclusion. That's the step most real-estate BI never takes. The dashboard shows you the dip; you're on your own for the cause. Being able to interrogate the dip in plain English — and get a sourced answer in the time it takes to read this sentence — is what turns a monitoring system into a decision-making one. ## How Travtus approaches this Travtus treats portfolio visibility as one layer of the [Everyday AI™ Platform](/everydayai) rather than a standalone reporting tool. Community Scores give you the daily, comparable health signal across every asset; [Scores](/scores) run 0-100 and refresh daily from operational signals. Community Profiles carry the context, [Reports](/reports) standardize what leadership reads, and Explore lets any user ask why a number moved and get a cited answer. It reasons across conversations, records, and everything you've published — no rip-and-replace of your existing systems. Operators like Cortland, MAA, and BH run this way today. The point isn't a prettier dashboard; it's that the intelligence layer is yours to build on top of the data you already have. Explore alone is delivering around a 15% productivity lift for the teams using it. For the wider view, see the [Data, Analytics & Portfolio Intelligence hub](/resources/category/data-analytics) and our take on why [clean data won't be enough in the age of AI](/resources/from-etl-to-intelligence-why-clean-data-wont-be-enough-in-the-age-of-ai). ## Frequently asked questions **How do I get real-time visibility across all my properties?** Score every asset on the same 0-100 scale, refreshed daily, and rank the portfolio worst-to-best. That turns fifty separate reports into one ordered list, so your attention goes to the assets that moved rather than the ones that didn't. **How do asset managers track portfolio risk in real time?** They watch leading operational signals — sentiment, maintenance backlog, renewal behavior — that move weeks before revenue. A daily score built from those signals flags a sliding asset early, while there's still time to change the result. **What is portfolio risk monitoring in real estate?** Continuous tracking of asset health across a portfolio using leading indicators, not just trailing financials, so risk surfaces while it's still cheap to fix rather than at quarter-end. **Why isn't monthly reporting enough?** Monthly reports describe what already happened; the operational cause is 30 to 90 days old by the time it lands. Real-time signals surface the cause while the outcome is still preventable. **Can I see why a property's health changed?** Yes — a score tells you an asset moved, and Explore tells you why, in plain language with citations drawn from conversations, records, and published reports. *See how daily [Scores](/scores) turn fifty reports into one ranked view of portfolio risk — [book a demo](/demo).* ## The Real ROI of AI in Property Management URL: https://www.travtus.com/resources/roi-of-ai-in-property-management Date: 2026-07-15 Summary: The honest ROI of AI in property management comes from three levers you can actually measure: labor hours returned, turnover avoided, and operational leakage closed — not a magic number. > **The real ROI of AI in property management comes from three measurable levers, not a magic number: labor hours returned from automating high-volume work, turnover avoided by responding faster and more consistently, and operational leakage closed on chargebacks, renewals, and slow turns. Add the three, compare to platform cost, and you get an honest payback — usually inside the first year, without a rip-and-replace.** It's the end of the quarter and your regional wants the number. Not a story about "efficiency" — the number. What did the software give us back? And if you've ever sat in that meeting, you know the honest answer is usually a shrug and a dashboard nobody trusts. That's the problem with most ROI-of-AI conversations in this business: they're built on vibes and vendor slides, not on the three things you can actually count. So let me give you the version I'd defend in that meeting. The return on AI in property management is real, but it isn't one big number that falls out of the sky. It's three levers, each measurable, that add up. Get relief on the work first, then do the math — because when you can show where the money comes from, the number holds up. ## How do you measure the ROI of AI in property management? Add three levers. That's the whole model. Every credible payback figure I've seen breaks down into labor hours returned, turnover avoided, and operational leakage closed. If a vendor can't tell you which of the three their number comes from, it isn't a number — it's a hope. Here's how each one works, and how to put a dollar on it. ## Lever one: labor hours returned This is the fastest and easiest to prove. You automate the high-volume, low-judgment work that buries your teams — intake, routing, reminders, status updates — and you measure the hours that used to go into it. The math is simple: hours automated per week, times your loaded hourly cost per person, times 52. Say a community manager spends ten hours a week just sorting and acknowledging inbound requests, and you automate eight of them. At a loaded cost of, say, $35 an hour, that's roughly $14,500 a year returned per person on that one workflow — and most portfolios have several. Operators on the Travtus platform typically see around a 15% productivity gain and close to 95% automation on the workflows they build, which is where this lever gets large. The reason this one shows up first: you can automate a well-scoped workflow quickly, and the hours are immediately visible. You don't wait two quarters to see it. It's usually enough on its own to cover the platform — the other two levers are upside. ## Lever two: turnover avoided This is the biggest line, and the one everyone underestimates. Every resident who moves out because they felt ignored costs you the full price of a turn: make-ready, marketing, the days the unit sits empty, and the staff time to re-lease it. Depending on your market, that's easily a few thousand dollars per door, sometimes far more. Residents don't leave because you sorted their email slowly. They leave because "ignored" adds up — the request that sat for a week, the follow-up that never came. When automation makes response times faster and more consistent, that specific friction goes away. Fewer people hit the point where they decide not to renew. To size it: take your annual move-outs attributable to service and responsiveness, estimate a realistic reduction, and multiply by your fully loaded cost per turn. Even a modest improvement moves real money, because the per-unit cost is high. Travtus operators have seen retention improvements near 25% where response times and follow-through tighten. Note what this lever is *not*: it has nothing to do with how you price or set rents. It's purely the cost of losing a resident you could have kept — a cost driver, not a pricing play. ## Lever three: operational leakage closed This is the quiet one. Leakage is money you were already owed or already spending that slips through the cracks of a busy, short-staffed operation: - **Chargebacks and billables** never captured because nobody had time to log them. - **Renewals** that lapsed because the follow-up fell through during a staffing crunch. - **Slow-turn vacancy days** — the extra days a unit sits empty because the turn workflow stalled between teams. - **Vendor and warranty recovery** that went unclaimed because the paperwork was buried. None of this is exotic. It's the stuff that goes missing when your team is triaging a 40-item pile every morning. When the routine coordination runs itself — reminders fire, handoffs happen, nothing waits on a human to remember — leakage closes. You can measure it directly: track captured chargebacks, on-time renewal follow-ups, and average turn days before and after. The delta is the lever. ## What's a realistic timeline — and what makes the number hold up? Be honest about sequencing, because that's where credibility lives. Labor hours returned show up in the first quarter — that's your early proof. Turnover avoided and leakage closed compound over the next two to four quarters as fewer move-outs and tighter follow-through work through the numbers. A well-scoped rollout typically reaches full payback inside the first year. Two things protect the number. First, no rip-and-replace. A platform that layers onto Yardi, RealPage, Entrata, or AppFolio avoids the migration cost and downtime that would otherwise eat your return before it starts. Owner-operators like Cortland, MAA, BH, Continental, and Preiss build on top of the system of record rather than tearing it out. Second, you build the workflows yourself, by describing them — so the cost of adding, adjusting, or expanding automation stays near zero. There's no developer backlog between you and the next hour saved. The work is yours to build and yours to change, which is exactly why the ROI keeps growing instead of plateauing after the initial project. ## How Travtus approaches this Travtus is built so the three ROI levers are things you can turn on and measure, not promises on a slide. The [Everyday AI™ Platform](/everydayai) runs a simple path — Connect to your existing systems, Explore your data, Create the primitives you need, Act — and you build the primitives (Reports, Profiles, Scores, and [Workflows](/workflows)) by describing them in plain language, on your own data, with your IP staying yours. That's what makes the payback honest: labor hours returned from automation, turnover avoided from faster response, and leakage closed from tighter coordination — all built on top of the system you already run, no rip-and-replace. Start with the highest-volume work, prove the hours, then compound the retention and leakage gains as you become an AI-native housing enterprise on one platform. For the fuller picture, see [Centralization](/centralization), the [Operations & Efficiency hub](/resources/category/operations-efficiency), and the deeper guides on [the ROI of AI in property management](/resources/roi-of-ai-in-property-management), [what to automate first](/resources/property-management-automation-what-to-automate-first), and [how to make a multifamily company AI-native](/resources/how-to-make-a-multifamily-company-ai-native). ## Frequently asked questions **How do you measure the ROI of AI in property management?** Measure three levers and add them up. First, labor hours returned: hours automated times loaded hourly cost. Second, turnover avoided: fewer resident move-outs times the full cost to turn and re-lease a unit. Third, operational leakage closed: chargebacks, renewals, and slow-turn vacancy days you were losing. Compare the total against platform cost for a real payback figure. **What is a realistic ROI timeline for AI in multifamily?** Labor hours returned show up first — often within the first quarter — because you can automate high-volume work quickly and measure the hours immediately. Retention and leakage gains compound over two to four quarters as fewer move-outs and closed leakage work through the numbers. A realistic full-payback window for a well-scoped rollout is inside the first year. **Does AI in property management require replacing your current systems?** No, and that matters for ROI. A platform that layers onto Yardi, RealPage, Entrata, or AppFolio avoids the cost, risk, and downtime of a rip-and-replace, so your return isn't eaten by a migration. You build on the system of record your teams already know and start measuring gains almost immediately. **What's the biggest hidden cost AI recovers in property management?** Turnover. Every avoided move-out saves the full cost of a unit turn plus lost days and re-leasing effort — often thousands of dollars per unit. Faster, more consistent response times keep residents from feeling ignored, which is a leading reason they leave. Retention is usually the largest and most underestimated line in the ROI model. **How much productivity gain can operators expect from AI?** Operators on the Travtus platform typically see around a 15% productivity gain and roughly 95% automation on the workflows they build, with retention improvements near 25% where response times and follow-through tighten. These are directional figures from real deployments, not guarantees — your numbers depend on which work you automate first. *Model your own three-lever payback with [Workflows](/workflows) you build by describing them — then [book a demo](/demo).* ## Is AI for Property Management Secure and Enterprise-Ready? URL: https://www.travtus.com/resources/secure-enterprise-ready-ai-property-management Date: 2026-07-15 Summary: The honest answer is that it depends on the platform. Here is how to tell whether AI for property management is genuinely secure, governable, and ready for your whole portfolio. > **AI for property management can be secure and enterprise-ready, but that depends on the platform, not on AI in the abstract. The signals to check are data ownership, role-based and property-based permissions, single sign-on, input screening before an agent acts, and explainable decisions. Get those right and security becomes an adoption enabler, not a blocker.** The most useful question I hear from operators is not "is AI safe?" It is "can we deploy this without exposing resident data, tripping a fair-housing issue, or losing control of decisions we cannot explain?" That is the right question, and the honest answer is that it depends far more on the platform you choose than on AI as a category. Security and governance are usually framed as the things that slow adoption down. In housing, I would argue the opposite. When the controls are strong and visible, teams stop hesitating. Legal signs off, IT signs off, and the rollout moves. The platforms that stall are the ones where nobody can answer a basic question about where the data went or why the system did what it did. ## Is AI for property management secure and compliant? It can be, and the difference lives in the architecture. Start with what "secure" should mean for a multifamily deployment: enterprise-grade infrastructure that is included rather than sold as an upgrade. On a secure AI platform multifamily teams can trust, data governance, role-based security, and property-based permissioning are built into every feature, not layered on for the enterprise tier. The practical controls to verify are concrete. Single sign-on via SAML 2.0 so access follows your existing identity rules. Role-based access so a leasing consultant and a regional manager see different things. Property-based permissioning so a user's reach is scoped to the assets they actually manage. Ask any vendor to show you these as defaults, not roadmap items. Compliance is related but not identical. A platform can be secure and still be configured in a way that creates fair-housing exposure. Security is about protecting data and access; compliance is about how the system is set up, supervised, and used in a regulated context. Both matter, and the second one is partly your responsibility, which is why I never treat any tool as compliant on its own. For specifics on your obligations, consult your own counsel; what follows are principles, not legal advice. ## How do I roll out AI across a property management company safely? The most common failure I see is not a security breach. It is the pilot that never becomes a program. A company turns on AI everywhere at once, the controls are inconsistent, the teams are overwhelmed, and the whole effort quietly stalls. The safer path is deliberately small at the start. Choose one clearly scoped use case on one property. Prove the outcome, document what the teams experienced, and only then expand to the next property and the next use case. This is the "yours to build" posture: you own the pace and the sequence, with the enterprise support there to lean on when you want it, rather than a big-bang rollout you cannot govern. What makes this expand-safely approach work is that the governance travels with you. When permissions and role-based security are built into every feature, widening the circle does not mean rebuilding controls from scratch each time. You are extending a governed system, not standing up a new one for every property. That is the difference between an experiment and an operating capability. ## What makes an AI platform enterprise-ready, not just functional? Enterprise-ready is less about features and more about defaults. The test I apply is simple: are security, governance, and permissioning present in every feature the moment you turn it on, or do you have to assemble them? Three things carry most of the weight. First, data ownership. You should own your data and your intellectual property. On Travtus that holds because everything is built on Connect, the governed data layer underneath the platform, so your information is not surrendered to an AI vendor as the price of using the product. Second, model independence. A model-agnostic platform means you are not locked to one AI provider's terms, pricing, or data handling; the platform can use the right model without your data being captive to it. Third, fit with what you already run. Enterprise-ready means no rip-and-replace: the platform works alongside Yardi, RealPage, Entrata, and AppFolio rather than demanding you tear them out. There is also a supervision layer worth naming. On Travtus, a Guardrail Agent classifies input for policy violations before an agent acts. That "before it acts" detail matters: it is the difference between catching a problem at the door and discovering it after a resident has already been affected. ## Can AI decisions in housing actually be explained and audited? This is the one I feel most strongly about. An AI decision you cannot explain is a liability, not an asset, and in housing that liability is not hypothetical. If a system influences how a resident is treated and no one can reconstruct why, you have a problem long before a regulator gets involved. Explainability is therefore not a nice-to-have; it is what lets your compliance function say yes. The controls to look for are a reviewable record of what an agent did and why, input screening before action, and permissions that make it clear who could have seen or triggered what. When those exist, an audit is a routine exercise rather than a fire drill. When they do not, every AI decision becomes an open question you cannot answer. Enablement belongs in the same conversation. Teams govern well only when they understand the system, which is why an Academy and documentation are part of readiness, not an afterthought. A control nobody understands is a control nobody uses. ## How Travtus approaches this Travtus treats security and governance as the foundation the product stands on, not a tier you buy up to. Enterprise-grade infrastructure, role-based security, and property-based permissioning are built into every feature; SSO uses SAML 2.0; the Guardrail Agent screens input for policy violations before an agent acts; and because the platform is built on Connect, you own your data and your IP. It is model-agnostic and requires no rip-and-replace of Yardi, RealPage, Entrata, or AppFolio, and operators including Cortland, MAA, BH, Continental, and Preiss are building on it. The invitation is straightforward: this is yours to build, at your pace, with us alongside you when you want the support. Start with one use case, prove it, and expand into an AI-native housing enterprise on one platform. Explore the [Everyday AI™ Platform](/everydayai), the controls behind it on [Security](/security) and [Compliance](/solutions/compliance), and the wider [Trust, Security & Compliance hub](/resources/category/trust-security-compliance). For the rollout mechanics, see [how to make a multifamily company AI-native](/resources/how-to-make-a-multifamily-company-ai-native) and [the ROI of AI in property management](/resources/roi-of-ai-in-property-management). ## Frequently asked questions **Is AI for property management secure and compliant?** It can be, but security and compliance depend on the platform, not on AI in general. Look for role-based security, property-based permissioning, SSO, input screening before an agent acts, and clear data ownership. Compliance also depends on how you configure and supervise the system, so treat these as controls you verify, not guarantees. **Does using AI mean handing my data to a third party?** Not on a well-designed platform. On Travtus, you own your data and your intellectual property because everything is built on Connect, your own governed data layer. A model-agnostic platform also means your data is not locked to a single AI vendor, so you keep control of where it lives and how it is used. **How do I roll out AI across a property management company without a failed pilot?** Start with one clearly scoped use case on one property, prove the outcome, then expand. This keeps risk contained, gives teams a real reference point, and avoids the all-at-once rollout that stalls at the pilot stage. Governance and permissions built into every feature let you widen the circle without rebuilding controls each time. **What makes an AI platform "enterprise-ready" for multifamily?** Enterprise-ready means the security, governance, and permissioning are part of every feature by default, not bolted on. Practically, that is SSO via SAML 2.0, role-based access, property-level permissions, auditability, explainable decisions, and integration with systems like Yardi, RealPage, Entrata, and AppFolio without a rip-and-replace. **Can AI decisions in housing be explained and audited?** They should be. An AI decision you cannot explain is a liability, especially in a regulated space like housing. Favor a platform that shows why an agent acted, screens input for policy violations before acting, and keeps a reviewable record. Explainability is what lets compliance teams sign off with confidence. *See the controls for yourself on [Security](/security), then [book a demo](/demo) to map a first use case.* ## How to Make Property Data Self-Service for Business Users URL: https://www.travtus.com/resources/self-service-property-data-for-business-users Date: 2026-07-15 Summary: The people who need the answer can't write the query, and the people who can write the query are three days behind. Here is how the best AI analytics platform for real estate data closes that gap. > **The best AI analytics platform for real estate data lets a business user ask a question in plain English and get a cited answer, without writing SQL or filing a ticket. It reasons across records, conversations, and published reports, and sits on your existing systems. That closes the gap between the person who has the question and the person who can query the data.** Here is the bottleneck I watched form at every organization I worked in. The regional manager has a question — which of my communities saw renewals soften last month? The answer exists; it's sitting in the data. But she can't write the query, so she files a request. The analyst who can write it is three days deep in a backlog. By the time the answer comes back, the meeting it was for is over and the question has changed. That's not a data problem. The data is fine. It's an access problem: the person with the question and the ability to write the query are two different people, and everything routes through the narrow one. Self-service analytics is the discipline of closing that gap — letting the person with the question get the answer directly. For years that meant training everyone in SQL or building dashboards for every conceivable question in advance. Neither worked. The real unlock is letting people ask in the language they already speak. ## How to make property data self-service for business users Remove the translation step. Today a business question has to be translated into a technical query by a technical person before the data can answer it. Every translation is a handoff, and every handoff is delay and lost context. Self-service works when that translation happens automatically. That's what a conversational AI layer does. Explore, inside Gateway, lets a user type the question the way they'd ask a colleague — "which communities had rising maintenance backlogs last month?" — and handles the translation itself. Under the hood it reasons across everything: it runs SQL against your structured records, vector search across unstructured text like resident conversations and notes, and web search where outside context helps. The user never sees any of that. They see a plain-language answer. The measure of success is simple. Count the questions in your organization that currently require a ticket to the data team. In most real-estate operations that's the overwhelming majority of day-to-day questions — occupancy trends, sentiment, maintenance status, renewal patterns. Self-service moves that population from "file a request and wait" to "ask and read." The teams using Explore this way are seeing roughly a 15% productivity lift, and that number is really just the recovered time that used to disappear into the request-and-wait cycle. ## Can I ask questions about my property data in plain English Yes — and the phrase "plain English" is doing real work here, because it's the difference between a tool for analysts and a tool for everyone. A dashboard makes you find the answer by navigating someone else's layout. A query language makes you know the schema. Plain-language ask makes neither demand. A worked example. A VP of operations wants to know why a specific community's health slipped. In the old model that's a ticket, a clarifying email, a query, and a two-day wait. With plain-language ask, she types "why did resident sentiment drop at [community] over the last six weeks?" and gets an answer that reasons across the resident conversations, the underlying records, and any published Reports or Scores for that asset — in the time it takes to read the response. The question and the answer live in the same minute. That's what makes it genuinely self-service: there's no second person in the loop. Crucially, this reaches across silos a dashboard can't. The answer to "why did sentiment drop" lives partly in structured survey scores and partly in the actual text of what residents wrote. A tool that only queries the database misses half of it. Reasoning across both — numbers and language together — is what lets a business user get a complete answer, not a partial one. ## What makes a good AI analytics platform for real estate data Three properties separate a platform you can actually roll out from a demo that impresses in a conference room: - **It reasons across everything.** Real-estate answers rarely live in one system. A platform that only reads structured records can't tell you why a number moved, because the why is usually in the conversations. Coverage across records, conversations, and published reports is non-negotiable. - **It answers with citations.** An answer without a source is a guess with good grammar. Citations let the user verify the result and let the data team audit what the tool is doing — which is what makes it safe to put in non-technical hands. Explore returns cited answers for exactly this reason. - **It sits on what you already have.** If adopting self-service requires migrating your systems first, it won't happen. The platform has to reason across your existing data with no rip-and-replace, so the path from "interested" to "asking questions" is measured in days, not a year-long implementation. Coverage, trust, and no migration. Miss any one and self-service stalls — either the answers are incomplete, or nobody trusts them, or you never get to production. ## How Travtus approaches this Self-service analytics is the front door to the [Everyday AI™ Platform](/everydayai). Explore lets any business user ask questions in plain language and get cited answers, reasoning across conversations, records, and everything you've published — and you build the underlying primitives, [Reports](/reports), Profiles, [Scores](/scores), and Workflows, simply by describing them, no engineering ticket required. It runs on your existing systems with no rip-and-replace, which is why operators like Cortland, MAA, and BH could put it in front of business users quickly. The result is [Customer Intelligence](/customer-intelligence) that isn't gated behind the data team — the answer is yours to build and yours to ask for. For more, see the [Data, Analytics & Portfolio Intelligence hub](/resources/category/data-analytics) and why [clean data won't be enough in the age of AI](/resources/from-etl-to-intelligence-why-clean-data-wont-be-enough-in-the-age-of-ai). ## Frequently asked questions **How do I make property data self-service for business users?** Give non-technical users a way to ask in plain English and get trustworthy answers, instead of routing every question through the data team. An AI layer translates the question into the right query and returns a cited answer. **Can I ask questions about my property data in plain English?** Yes. With Explore you type the question the way you'd ask a colleague, and it reasons across records, conversations, and published reports to answer in plain language with citations. **What makes a good AI analytics platform for real estate data?** It reasons across everything, answers with citations, and sits on your existing systems without a rip-and-replace. Coverage, trust, and no migration. **Does self-service replace the data team?** No — it reassigns them. Routine lookups get answered directly, so analysts move to modeling, governance, and the hard analysis only they can do. **How do I trust an AI answer?** Insist on citations. A platform that shows which records, conversations, or reports it drew from lets users verify and lets the data team audit — which is what makes self-service safe to scale. *Put [Customer Intelligence](/customer-intelligence) in the hands of the people asking the questions — [book a demo](/demo).* ## The State of Multifamily AI in 2026: Yours to Build URL: https://www.travtus.com/resources/state-of-multifamily-ai-2026 Date: 2026-07-15 Summary: A July 2026 snapshot of multifamily AI: two camps have emerged — stalled point-tool pilots and AI-native leaders operating on one platform — and the gap between them is widening fast. > **In 2026 multifamily AI has split into two camps: operators whose point-tool pilots stalled at a single task, and AI-native leaders running one governed platform connected to the systems they already own. The second camp is pulling ahead because building on your own data — describing what you want rather than coding it — compounds, and IP stays yours.** I spent twelve years building enterprise software before proptech, and I have watched this exact pattern before: a wave of narrow tools arrives, buyers accumulate a dozen of them, and then someone asks the uncomfortable question of what actually changed. Multifamily is at that moment now. The honest read on July 2026 is not that AI failed in housing — it is that two very different bets are now producing very different results. This is a market snapshot, not a forecast dressed up as one. Here is where things actually stand, what changed to get us here, and where the next twelve months point. ## What is the state of AI in multifamily in 2026? The state of the market in 2026 is a clean split into two camps. The first and larger camp bought point tools. A leasing bot here, a review-response tool there, a maintenance triage add-on somewhere else. Each one automated a single task, each held its own slice of resident and property data, and each arrived with its own contract and login. These pilots often worked in the narrow sense — the demo did what it promised — and then stalled. They never spread past the team that bought them, and they never changed how the company operated. An operator I spoke with this spring had eleven separate AI experiments running and could not name one that had reached the whole portfolio. The second, smaller camp went AI-native. Instead of buying tasks, they adopted one platform, connected it to the property management systems they already ran, and started building their own working pieces on top of their own data. That group is now compounding: every report, profile, score, or workflow they build makes the next one cheaper. The gap between the two camps is the real headline of 2026, and it is widening. ## What is multifamily AI, and what changed? Multifamily AI is the use of artificial intelligence to run rental-housing operations — understanding resident conversations, surfacing portfolio signal, and automating repetitive work across leasing, maintenance, and resident service. That definition has not changed. What changed is the delivery model, and three shifts did most of the work. - **Data got connected instead of copied.** The early tools each demanded their own export. The platforms read directly from Yardi, RealPage, Entrata, and AppFolio, so there was one connected picture instead of a dozen partial ones. - **Building stopped meaning coding.** In the AI-native camp, an operations leader now builds a workflow by describing what they want in plain language — not by filing an IT ticket and waiting a quarter. That single change is why the logic finally sits with the people who understand the work. - **Governance moved into the product.** Enterprise-grade governance stopped being a security review bolted on at the end and became part of every feature. That is what let AI move from a fenced-off pilot to something a compliance-minded enterprise could run across a whole portfolio. The proof is not hypothetical. Operators including Cortland, MAA, BH, Continental, and Preiss have moved down the platform path rather than accumulating tools, and the pattern in the AI-native camp is consistent: roughly 95% automation on the workflows they build, around 15% productivity gains where conversational insight replaces manual digging, and retention improvements in the range of 25% where service quality scales without adding headcount. Those are the numbers behind the split — and the reason the [platform-versus-point-solution](/resources/platform-vs-point-solutions) question stopped being academic this year. ## What does the adoption data show? The two-camp read stopped being anecdote this year — the industry's own survey data now shows both the surge and the split: - **Adoption is surging.** 34% of property management professionals used AI in 2025, up from 21% a year earlier, per the [AppFolio Benchmark Report](https://naahq.org/flat-rents-ai-adoption-2025-appfolio-property-management-benchmark-report-reveals-key-trends). By mid-2026, [89% of multifamily operators have introduced AI in some form](https://www.globenewswire.com/news-release/2026/07/29/3335276/0/en/EliseAI-Unveils-New-State-of-AI-in-Multifamily-Report-85-See-Reduced-Operating-Expenses.html). - **But embedding is rare.** In that same 2026 survey, only 34% have fully embedded AI into daily operations. The distance between "introduced" and "embedded" is the two camps expressed as a single number. - **The split follows scale.** [47% of operators managing more than 5,000 units use AI, against 28% of the smallest operators](https://www.bisnow.com/national/news/top-talent/apartments-ai-driven-revolution-creating-tech-haves-and-havenots-128182) — the enterprise camp is pulling away. - **Expectations have diverged with it.** Firms implementing AI expect [31% portfolio growth versus 12% for firms without it](https://www.appfolio.com/newsroom/property-manager-benchmark-survey-2026). - **The operating pressure isn't easing.** Onsite turnover hit [29.2% in 2025](https://naahq.org/news/why-employee-retention-challenges-go-deeper-wages), and [66% of property management leaders' time goes to routine and reactive work](https://www.appfolio.com/blog/naa-top-challenges-2025) — the labor math that makes automation non-optional. - **Capital has picked a side.** [$16.7 billion went into proptech in 2025, up 67.9% year over year](https://www.multifamilydive.com/news/proptech-investment-venture-capital-funding/809517/), with investors explicitly favoring AI-enabled solutions built on strong data foundations. Read together, the data says what the two camps already know: nearly everyone now *has* AI, few *run on* it, and the operators who crossed that line are compounding. Crossing it is a platform decision, not a purchasing spree — which is the argument of this whole piece. ## Where does multifamily AI go from here? The direction of travel is clear, even if the pace varies by operator. First, consolidation. The dozen-contract sprawl is expensive to run and impossible to govern, and finance teams have noticed. Expect operators to collapse experiments onto a smaller number of platforms that carry governance natively. The [clean-slate advantage](/resources/the-clean-slate-advantage-why-late-adopters-may-win-the-ai-race) is real here — operators who stalled in the first wave can skip straight to a platform without unwinding much. Second, the center of gravity moves to the business user. The competitive edge in 2027 will not belong to whoever licensed the most tools; it will belong to whoever can turn their own data and their own people into working automation fastest. When building means describing what you want, the operations and service leaders who know the work become the builders — and the IP they create stays with the company. Third, the definition of "AI-native" hardens into an operating posture rather than a feature list. Being AI-native is not owning AI features; it is running the enterprise so that everyday work is understood, surfaced, and automated by default, on one governed platform. That is the bar the leading camp is setting, and it is the one the rest of the market will be measured against. ## How Travtus approaches this Travtus built the [Everyday AI™ Platform](/everydayai) for the AI-native camp — and to give the stalled camp a way to join it without starting over. The model is a connected loop: **Connect → Explore → Create → Act.** Connect reads the systems you already run — Yardi, RealPage, Entrata, AppFolio — with no rip-and-replace. [Explore](/explore) lets your teams ask questions of that data in plain language. Create is where they build the primitives that matter — Reports, Profiles, Scores, and Workflows — by describing what they want, not by writing code. Act puts those builds to work across the portfolio. Enterprise-grade governance is carried in every feature, not added at the end, which is what makes the platform safe to run at scale rather than in a fenced-off pilot. The goal is straightforward and it is the one the market is converging on: become an AI-native housing enterprise on one platform. The AI-native operation is yours to build — on your data, by your people, with your IP staying yours. For the wider picture, the [Everyday AI & the AI-Native Operator hub](/resources/category/everyday-ai) collects the rest of this thinking, and the [platform overview](/platform) shows how the loop fits together. ## Frequently asked questions **What is the state of AI in multifamily in 2026?** The market has split into two camps: operators whose point-tool pilots automated one task and stalled, and AI-native leaders running one governed platform on their existing systems. The second camp is pulling clearly ahead. **What is multifamily AI?** It is the use of AI to run rental-housing operations — understanding resident conversations, surfacing portfolio insight, and automating repetitive work. In 2026 the meaningful version is platform-based and connected to the systems you already run. **Why did so many pilots stall?** Each point tool automated a narrow task, held its own data, and never changed how the company operated, leaving buyers with a dozen experiments and no compounding value. **Do I have to replace my PMS?** No. The leading operators kept Yardi, RealPage, Entrata, or AppFolio and connected a platform on top — there is no rip-and-replace. **How many operators use AI in 2026?** Surveys show 89% have introduced AI in some form, but only about 34% have embedded it into daily operations — and adoption splits by scale, 47% of 5,000+-unit operators versus 28% of the smallest. **Where does it go next?** Toward consolidation on governed platforms and toward business users building their own AI logic by describing what they want. *See how the connected loop works on the [Everyday AI™ Platform](/platform), or [book a demo](/demo) to map it to your portfolio.* ## How to Underwrite a Multifamily Deal Beyond the Financials URL: https://www.travtus.com/resources/underwrite-multifamily-beyond-financials Date: 2026-07-15 Summary: In-place NOI tells you what the asset earned, not whether it will keep earning. Here is how to improve underwriting accuracy in multifamily by reading the operational signals financials hide. > **To improve underwriting accuracy in multifamily, underwrite the operation, not just the income statement. In-place NOI is a trailing number; whether it holds depends on resident sentiment, maintenance backlog, and retention — leading signals that move months before revenue. Reading them tells you if you're buying durable income or a T-12 propped up by deferred work and departing residents.** Every underwriting model I've ever built starts with the trailing twelve. It's the right anchor — audited, comparable, real. But the T-12 answers only one question: what did this asset earn? The question that actually determines your return is a different one: will it keep earning that once the current operator hands you the keys? The financials are silent on that, and the gap between the two is where deals go wrong. I learned this on the asset-management side, inheriting communities that underwrote clean and then missed year-one by a wide margin. The income was real; the operation underneath it wasn't. The seller had earned that NOI by deferring maintenance and pushing renewals that were never going to hold. The numbers didn't lie — they just didn't tell me the whole truth, because the whole truth wasn't in the numbers yet. ## How to improve underwriting accuracy in multifamily Add a layer that tests the durability of in-place NOI. Financial due diligence tells you the income is real; operational due diligence tells you whether it's repeatable. Those are different diligence questions, and most models only ask the first. Consider two deals. Both show $3.0M in-place NOI, both at 94% occupancy, both with clean T-12s. On a spreadsheet they underwrite identically, and you'd price them the same. Now add the operational read. Asset A has flat resident sentiment, a maintenance backlog under control, and renewals holding at 55%. Asset B has sentiment trending negative, a backlog up sharply over the trailing months, and renewals that have quietly slipped to 40% — propped up short-term by aggressive concessions. Same NOI today. But Asset B is going to hand you elevated turn cost, a capital catch-up on deferred repairs, and an occupancy dip in year one. The financials rated them equal; the operation says one is meaningfully riskier. That's the accuracy you're missing when you stop at the income statement. The correction isn't to distrust the T-12. It's to underwrite the operation that produces it, so your first-year assumptions rest on how the asset actually runs. ## How to assess the operational health of an apartment community before buying Three signals do most of the work, because each leads a different part of NOI: - **Resident sentiment.** How residents talk about the community — in messages, reviews, survey responses — leads renewal behavior by weeks to months. A community where sentiment is deteriorating is telling you where next year's move-outs are coming from, before they show up as vacancy. If you can read the tone of the resident base during diligence, you're reading a leading indicator of retention. - **Maintenance backlog.** The size and age of the open work-order backlog is the clearest read on deferred cost. A backlog that's been growing is capital and turn expense the seller pushed into your hold period. It also predicts dissatisfaction, which loops back into sentiment and renewals. A T-12 with suspiciously low repair spend and a large aged backlog isn't efficiency — it's a deferred bill. - **Retention.** Renewal rate and trend are the most direct link to future occupancy and revenue. Two assets at the same occupancy but with 55% versus 40% renewal are on very different trajectories; the second is running harder just to stand still, and that shows up as turn cost and concession pressure the moment you take over. Read together, these three tell you whether the seller's NOI is the product of a healthy operation or a fragile one. That's the operational health picture the financials can't give you — and it's exactly the picture that determines your first-year performance. ## Why operational due diligence rarely happens at deal speed The reason underwriters lean on financials isn't that they think operations don't matter. It's that operational data is hard to get to inside a deal timeline. Sentiment lives in review sites and message logs, maintenance in a work-order system, renewals in the property management platform — three systems, three exports, and days of analyst time to reconcile per deal. When you're screening dozens of opportunities, that work simply doesn't get done, so the model defaults to the financials it can pull quickly. That constraint is what's changed. When an AI layer can reason across conversations, records, and published reports at once, you can ask the operational questions directly instead of building the analysis by hand. "Summarize resident sentiment trend at this community over the last six months." "How large is the maintenance backlog and how has it aged?" "What's the renewal rate and is it holding?" Answers come back in plain language, with citations, in minutes rather than days. Operational due diligence stops being a luxury reserved for the one deal you've already fallen in love with and becomes something you can run on everything in the pipeline. That's the shift that makes beyond-the-financials underwriting practical at portfolio scale — and it's a diligence edge that's yours to build into your own process, not a black box you have to trust. ## How Travtus approaches this Travtus surfaces the operational signals that predict whether NOI is durable, as part of the [Everyday AI™ Platform](/everydayai). Community and Customer [Scores](/scores) run 0-100 and refresh daily, giving you a normalized read on asset health you can use in diligence as readily as in asset management. [Reports](/reports) standardize the operational picture, and Explore lets acquisitions and asset-management teams ask sentiment, maintenance, and retention questions in plain language and get cited answers — reasoning across conversations, records, and everything published, using SQL, vector, and web search. There's no rip-and-replace; the intelligence layer sits on the data you already have. Operators including Cortland, MAA, and Continental use it to connect the operation to the number. For the strategic frame, see [Asset Management](/solutions/asset-management), how AI [boosts apartment operations and financial performance](/resources/using-ai-to-boost-apartment-operations-and-financial-performance), and the [Data, Analytics & Portfolio Intelligence hub](/resources/category/data-analytics). ## Frequently asked questions **How do I improve underwriting accuracy in multifamily?** Test whether in-place NOI is durable, not just what it is. Layer operational due diligence — sentiment, maintenance backlog, retention — onto the financials to see whether current income rests on a stable operation or a fragile one. **How do I assess operational health before buying?** Read three signals past the T-12: how residents talk about the community, how large and aged the maintenance backlog is, and whether renewals are holding. Together they reveal whether the NOI is well-run income or deferred cost and churn. **Why can two deals with identical NOI carry different risk?** Because identical income can sit on very different operations — one earned with satisfied residents and current maintenance, the other by deferring repairs and pushing renewals that won't stick. **What operational signals predict future NOI?** Sentiment leads renewals, maintenance backlog leads dissatisfaction and turn cost, and retention leads occupancy — all moving weeks to months before the income statement. **Can operational due diligence be done at scale?** Yes. An AI layer that reasons across sentiment, work-order, and renewal data lets you ask the questions in plain language and get cited answers in minutes, so diligence scales with deal flow. *Underwrite the operation behind the NOI with [Asset Management](/solutions/asset-management) intelligence — [book a demo](/demo).* ## What is Agentic AI for Multifamily Operators? URL: https://www.travtus.com/resources/agentic-ai-multifamily-operators Date: 2026-04-01 Summary: Most AI in multifamily still responds to prompts or automates tasks. Agentic AI changes that, coordinating actions across systems and reshaping how work gets done. Here is what that means in practice. Artificial intelligence in multifamily has, for most of its history, meant one of two things. A tool that answers a specific question. Or a feature that automates a specific task. Both are useful. Neither is what people mean when they talk about agentic AI. **Agentic AI is AI that can take actions across systems to achieve a goal, rather than simply respond to a prompt.** Agentic AI represents a meaningful shift in how AI operates within an organization. Understanding what it is, and what it makes possible, is becoming increasingly relevant for multifamily operators thinking about the next phase of their technology strategy. ## The Difference Between AI Features and Agentic AI Most AI in multifamily today is reactive and narrow. It operates within the boundaries of a single tool and is designed to perform a specific task. Agentic AI works differently. It operates continuously, taking sequences of actions, making decisions across multiple steps, and working toward broader goals with limited human intervention at each stage. In practical terms, this means an agentic system can move from identifying an issue to gathering context to triggering the appropriate response, without a team member having to coordinate each step manually. This is not automation in the traditional sense. It is a more capable form of intelligence that can handle complexity across workflows, systems and teams. ## Why It Matters for Multifamily Operations Multifamily operations involve a high volume of recurring, interconnected tasks. Leasing follow-up. Maintenance coordination. Resident communications. Performance reporting. Compliance tracking. In most organizations, these tasks are handled by people working across multiple systems, often without a shared view of the information they need. The work gets done, but it requires significant coordination and leaves limited capacity for higher-value activity. Agentic AI changes the economics of this model. When AI can handle multi-step workflows autonomously, coordinating information across systems and triggering actions based on real-time data, teams are freed from the coordination overhead that currently consumes so much of their time. The result is not just efficiency. It is a shift in what teams are able to focus on. ## What Agentic AI Looks Like in Practice For multifamily operators, agentic AI is most valuable when it is embedded into the workflows that run every day. Consider portfolio oversight. Rather than waiting for a weekly report, an agentic system can continuously monitor performance data across communities, surface early indicators of risk, and flag the situations that require attention before they become material issues. Or consider a leasing workflow. An agentic system can monitor incoming leads, assess intent based on conversation history, prioritise follow-up based on a prospect's profile, trigger the appropriate communication, and update the relevant records across systems. Each step follows from the last, with the system adapting based on what it learns along the way. In both cases, the value is not in any single action. It is in the ability to coordinate multiple actions intelligently, across systems and over time. This is closely related to how AI shows up in day-to-day operations, which we explore in more detail in our post on [**Everyday AI in Property Management.**](/resources/everyday-ai-property-management) ## The Foundation Agentic AI Requires Agentic AI does not operate in isolation. Its effectiveness depends on the quality and connectivity of the data it can access. This is where many organizations find that their current technology stack becomes a constraint. Data is distributed across systems. Context is fragmented. Workflows are disconnected. Agentic systems require a different foundation. One where data from across property management systems, communication tools and operational platforms is connected and accessible in real time. Without this connected intelligence layer, AI remains limited to individual tools and isolated use cases. Building toward agentic AI means investing in that foundation. Not just adding AI features, but enabling intelligence to operate across the organization. ## Where Multifamily Is Heading Agentic AI is not a distant development. Operators investing in connected data infrastructure today are building the foundation for what comes next. As a result, many are rethinking their approach to technology, moving away from disconnected point solutions toward [**platforms that can support coordinated intelligence and action.**](/resources/platform-vs-point-solutions) The organizations that move earliest are likely to see the most significant operational advantage, not because the technology is difficult to understand, but because the underlying foundation takes time to build. For senior stakeholders evaluating their technology strategy, the relevant question is not whether agentic AI will become important in multifamily operations. It is whether the foundation to support it is already in place. ## Travtus and Agentic AI Travtus is an intelligence platform built for multifamily housing operators. The Everyday(AI)™ platform connects data across systems, creating the foundation required for agentic workflows to operate effectively. It enables teams to move from insight to action by coordinating decisions and automation across systems and teams. This allows organizations to embed intelligence into everyday operations, not as a separate tool, but as part of how work gets done. ### [**\[Request a demo\]**](/demo) ## AI Platform vs Point Solutions for Multifamily URL: https://www.travtus.com/resources/platform-vs-point-solutions Date: 2026-04-01 Summary: Point AI tools win quick pilots and stall at scale — sprawl, integration debt, siloed context. Here is why operators are consolidating onto one platform, when a point tool is still the right call, and a four-test framework for choosing the platform layer. > **A point solution automates one task in one department; an AI platform connects your data, intelligence and workflows across the whole business on one foundation. Multifamily operators are consolidating because adding more point tools multiplied the visibility problem instead of solving it — and a platform lets every new use case run on the same connected data, with no rip-and-replace.** Here is the buying pattern we see most often in enterprise multifamily right now: a leasing AI bought in 2024, a call-center AI added in 2025, a collections pilot running somewhere in the portfolio, and a renewals tool in procurement. Four vendors, four contracts, four security reviews — and none of them can see what the others know. The question executives are asking is no longer "should we use AI?" It's "why do we own this much AI and still not operate differently?" That's the platform-versus-point-solution decision, and it's worth making deliberately rather than by accumulation. ## What's the difference between an AI platform and a point solution? A **point solution** is built to do one job: answer leasing inquiries, chase delinquency, schedule maintenance. It's fast to buy, fast to pilot, and its ceiling is exactly the job it was built for. An **AI platform** is infrastructure. It connects to the systems you already run, builds context across departments — resident history, maintenance load, team processes, portfolio data — and lets your teams deploy AI against any use case on that shared foundation. The structural difference is **context**. A leasing bot knows the conversation in front of it. A platform knows the resident has two open work orders, went through a rough move-in, and is 60 days from renewal — and every agent, workflow, and report on the platform shares that knowledge. In operations, context is the difference between an answer and the right answer. | Capability | Point Solutions | AI Platform | | --- | --- | --- | | Scope | One task or department | Cross-departmental | | Data access | Distributed across tools | Unified across systems | | Integration | Managed tool by tool | Connected at the source | | Context | Sees only its own channel | Shared across systems and teams | | Who builds on it | The vendor's roadmap | Your business teams | | Leadership visibility | Dependent on reports | Real-time and accessible | | Scalability | A new integration for each use case | New use cases on the same foundation | ## Why do point AI solutions stall in multifamily? Three failure modes show up consistently: 1. **Pilot purgatory.** The tool works at five properties and never reaches five hundred. Point tools are bought by one department for one problem, so nobody owns scaling them across the enterprise — and the business case was never built for it. 2. **Sprawl and its overhead.** Each tool is its own contract, security review, integration, admin console, and training burden. IT ends up configuring a growing shelf of narrow tools while the business waits. The spend rises linearly; the capability doesn't. 3. **Siloed context makes every tool dumber.** Your leasing AI, resident-comms AI, and collections AI each hold a fragment of the same resident's story and share none of it. Residents feel it — they repeat themselves to your own systems. Leaders feel it — insight stays trapped in per-tool dashboards, so there is no single view of what's actually happening across the portfolio. None of this means point tools are bad software. It means single-task tools have a structural ceiling in a business where the work is connected. ## When is a point solution the right choice? Honest answer: sometimes. A point solution is defensible when the task is genuinely isolated — narrow scope, no shared-context benefit, no ambition to expand. If you operate a small portfolio and want to answer after-hours leasing calls, a focused tool may be all you need. The calculus flips when any of these are true: - You expect to run AI in **more than one department** within two years. - The use cases **share data or residents** (in multifamily, they almost always do). - You're an **enterprise operator** where procurement, security, and change management make each new vendor expensive. - You want your **own teams** to configure how AI behaves, instead of waiting on each vendor's roadmap. For REITs and PE-backed owner-operators, that describes essentially every AI decision on the table. ## What's the best alternative to point AI solutions for real estate? An enterprise AI platform that meets four tests: 1. **It integrates with your existing stack — no rip-and-replace.** It should sit on top of Yardi, RealPage, Entrata, or AppFolio, not compete with them. Your PMS keeps managing leases and financials; the platform adds the intelligence layer. 2. **It carries context across departments.** The same operational understanding should power resident communications, maintenance, retention signals, and reporting — that's where the compounding comes from. 3. **Your business users can build on it.** Operations teams should author workflows, reports, and process guides themselves. If every change is an IT ticket or a vendor request, you've bought a bigger point solution. 4. **It's governed like enterprise infrastructure.** One security review, one audit surface, one place to control what AI can do — instead of a dozen shadow deployments. Run the platform decision the way you'd run any infrastructure decision — not "which demo was best?" but "which foundation will every future use case stand on?" ## How Travtus approaches this Travtus is built as that foundation: the [Everyday AI™ Platform](/everydayai) for housing operators. It connects to the systems you already rely on and makes that data usable across teams — AI agents handling resident conversations, workflows your operations team authors without engineering, plain-language access to your own operating data through Explore, and scores and reports that turn everyday interactions into portfolio intelligence. Operators building workflows on the platform report roughly 95% automation on the specific workflows they build, and teams using Explore see around a 15% productivity gain from easier access to information. One platform, one security review, one context layer — and each new use case makes the last one smarter. As that foundation matures, it also enables more advanced forms of AI that coordinate actions across systems — we explore that direction in [Agentic AI for Multifamily Operators](/resources/agentic-ai-multifamily-operators). For the category basics, start with [What is multifamily AI?](/resources/what-is-multifamily-ai); if sprawl is already the problem, see [How to reduce AI vendor sprawl in real estate](/resources/reduce-ai-vendor-sprawl-real-estate). More in [Everyday AI & the AI-Native Operator](/resources/category/everyday-ai). You do not need to replace your existing technology. You can connect it. ## Frequently asked questions **What is the difference between an AI platform and point solutions?** A point solution automates one task inside one department; an AI platform connects data across all your systems into one governed foundation where intelligence and workflows are shared. Point tools add capability tool by tool; a platform compounds, because every new use case runs on the same connected data instead of a new integration. **Why do point AI solutions stall in multifamily?** Three failure modes recur: pilot purgatory, where a tool works at five properties but nobody owns scaling it to five hundred; vendor sprawl, where each tool adds its own contract, security review, and integration; and siloed context — each AI holds a fragment of the same resident's story and shares none of it. **When is a point solution the right choice?** When the task is genuinely isolated — narrow scope, no shared-context benefit, no plan to expand AI into other departments. The calculus flips once use cases share data or residents, or once procurement and security overhead make each new vendor expensive. **What is the best alternative to point AI solutions for real estate?** An enterprise AI platform that integrates with your existing PMS, carries context across departments, lets business users build on it without engineering, and is governed centrally like infrastructure. **Do I have to replace my PMS to adopt an AI platform?** No. A platform like Travtus connects to Yardi, RealPage, Entrata or AppFolio — your PMS keeps managing leases and financials while the platform adds the intelligence and automation layer on top. --- *Weighing a specific tool against a platform approach? [Book a demo](/demo) — bring your actual use-case list.* ## What is Everyday AI in Property Management? URL: https://www.travtus.com/resources/everyday-ai-property-management Date: 2026-03-10 Summary: Everyday AI embeds intelligence into day-to-day operations, making it available where decisions are made and actions are taken. The result is more consistent execution across teams and across the portfolio. Most conversations about AI in property management focus on what AI can do. Capabilities, features and use cases. Less attention gets paid to a more practical question: how does AI actually fit into the way teams work every day? This matters more than it might seem. AI that requires teams to change how they work in order to use it tends to get used inconsistently or not at all. The tools get adopted, the dashboards get built and six months later the same teams are operating in much the same way they always did. Everyday(AI)™ is a different approach. **Everyday AI is part of how teams work day to day, not something they have to access separately.** It also reflects a broader shift in how AI operates, [**moving toward systems that can act and coordinate across processes**](/resources/agentic-ai-multifamily-operators), not just respond to inputs. The principle is straightforward: AI should operate as part of normal operations, not sit alongside them as a separate system that teams have to go and find. ## The Gap Between AI Capability and AI Adoption There is a gap in most multifamily organizations between what their AI tools are capable of and how consistently those tools are used. Part of this is a training and change management challenge. But a larger part is structural. When AI is delivered through a dedicated dashboard, a separate platform or a scheduled report, it asks teams to step outside their normal way of working in order to access it. For teams managing high volumes of daily tasks, that friction is significant. The insight arrives too late or gets missed entirely. The tool that was supposed to improve decision-making ends up being consulted only when someone remembers to check it. Everyday AI addresses this by making intelligence available as part of how teams already operate, rather than adding a new place they need to visit. ## What Everyday AI Looks Like in Practice In a property management context, Everyday AI means intelligence and automation are present at the point where decisions get made and actions get taken. For an onsite leasing team, For an onsite leasing team, this means relevant prospect context, follow-up priorities and communication history are surfaced, with the next actions triggered automatically. For a central operations team, it means performance signals and emerging issues across the portfolio are visible in real time without waiting for a weekly summary. For senior leadership, it means portfolio-level intelligence is accessible when it is needed, in plain language without requiring a data analyst to prepare it. In each case, the AI is present in the flow of work rather than sitting separately from it. The result is more consistent use, faster decisions and less time spent retrieving information that should already be available. ## The Role of Connected Data Everyday AI is only possible when the underlying data is connected. If intelligence is to be available at the point of decision, the data that powers it needs to be accessible across systems in real time. Leasing data, resident communications, maintenance records, financial performance and operational metrics all need to be available from a shared foundation. This is where the architecture of the technology stack becomes relevant. Point solutions that hold data independently cannot support this model. A connected platform that unifies data across systems can. This is explored further in [**our comparison of platforms and point solutions in multifamily technology.**](/resources/platform-vs-point-solutions) For multifamily operators, building toward Everyday AI means investing in data connectivity first. The intelligence layer depends on it. ## Consistency at Scale One of the less obvious benefits of Everyday AI is what it does for operational consistency across large portfolios. When AI is part of how teams work day to day, the quality of decisions and actions becomes less dependent on individual experience or availability. A leasing agent with six months of experience is working with the same contextual intelligence as one with six years. An onsite team at a smaller community has access to the same operational insight as one at a flagship property. This matters at scale. For operators managing hundreds or thousands of units across multiple markets, consistency is one of the hardest things to maintain. Everyday AI does not solve every consistency challenge, but it does raise the floor across the organization. ## Moving from Periodic Insight to Continuous Intelligence The shift toward Everyday AI represents a broader change in how multifamily operators think about intelligence in their organizations. Periodic insight, delivered through reports and dashboards, along with disconnected automations running in individual tools, was the model that made sense when AI was a specialist capability. Continuous intelligence and coordinated automation, available as part of day-to-day operations, is the model that makes sense now. This is not just about understanding what is happening. It is about acting on it. Intelligence surfaces what needs attention, clarifies the actions required and enables those actions to be carried out automatically across systems. This only works when built on a shared foundation. When data, intelligence and automation are connected, new use cases can be introduced without adding new tools or rebuilding integrations. That ease of scalability is what drives adoption. When new capabilities can be introduced without changing how teams work, AI becomes part of day-to-day operations rather than something that needs to be rolled out each time. For senior stakeholders evaluating their technology strategy, the question is not just what AI tools to adopt. It is how to build an organization where intelligence is part of how work gets done, not something that gets consulted occasionally when the right person remembers to look. That is what Everyday AI makes possible. ## Travtus and Everyday AI Travtus is an intelligence platform built for multifamily housing operators. The Everyday(AI)™ platform is designed around this principle. It connects data across systems, creates the intelligence layer that allows teams to explore operational context, trigger automation and act on real-time insight as part of daily operations. This is not a separate tool that teams need to consult. It is intelligence embedded into how work gets done across leasing, operations and leadership every day. ### [**\[Request a demo\]**](/demo) ## Leading in the Age of AI: A Guide for Business Leaders URL: https://www.travtus.com/resources/ai-guide Date: 2025-12-09 Summary: Explore how AI evolved from data-heavy systems to advanced models and agents. Learn the shift from AI 1.0 to AI 2.0, how LLMs work, and why leaders need AI literacy to navigate technology, improve operations, and drive strategic value ### A Decade of Transformation: AI from 2016 to 2026 **2016 – 2019: ****Early Momentum — AI 1.0** This was AI 1.0 — when progress depended on structured data, data scientists, and hosted models. Success required specialized teams, complex infrastructure, and clearly labeled datasets. - **Image and voice breakthroughs**: AI models like ResNet (2015) (for images) and WaveNet (for voice) made machines “see” and “talk” better. - **Chatbots began**: Basic customer service bots appeared but were limited and scripted. - **Business use**: Early automation in finance, logistics, and customer service **2020 – 2023: ****The Rise of Large Language Models (LLMs) — AI 2.0 Emerges** AI 2.0 began to emerge, marking a shift from narrow, data-specific systems to generalized models that could adapt to new contexts. The pace of progress accelerated dramatically — models improved in months, not years. - **GPT-3 (2020)** marked a major shift in generative AI: It could write emails, articles, even code. - **BERT (2018) & Transformers (2017)**: Helped models understand context, not just keywords. - **Widespread adoption**: AI started helping with document analysis, customer communication, and personalized marketing. **2024 – 2026: ****Intelligence at Work — AI 2.0 in Practice** AI 2.0 is now operational. Prompting, fine-tuning, and agentic behavior have become part of daily workflows. New roles like prompt engineers bridge human expertise with machine intelligence. - **Multimodal AI**: Systems now understand and generate text, images, video, and voice together. - **AI agents**: Not just answering questions, but taking action (e.g., booking, emailing, summarizing). - **Business transformation**: AI copilots are part of sales, HR, legal, ops — acting as “virtual team members.” * * * #### AI 1.0 vs AI 2.0: How We Got Here Over the past decade, AI has evolved from data science projects to intelligent, adaptive systems embedded into everyday work. | | Focus | How It Worked | Who Drove It | Limitations | | --- | --- | --- | --- | --- | | AI 1.0
2016 – 2021 | Prediction & classification | Models trained on labeled data for narrow tasks | Data scientists, engineers | Complex setup, costly hosting, siloed insights | | AI 2.0
2022 – 2026 | Generation & reasoning | Foundation models that create, converse & act | Cross functional engineering teams, data pipelines | Still maturing, needs governance & context | **In AI 1.0,** success meant building models — training on structured datasets and deploying them through technical teams. **In AI 2.0,** success means orchestrating intelligence — using powerful, pretrained models, prompting them in natural language, and orchestrating them across workflows using smarter pipelines and data architecture. ### Why Leaders need to understand AI AI is no longer just a technical topic — it’s a strategic one. The next generation of competitive advantage depends on how leaders deploy, govern, and scale AI across their organizations. Understanding AI allows leaders to:
Make better decisionsKnow what’s possible (and what’s hype). Recognize which tasks or processes AI can realistically automate or enhance.
Shape culture and capabilityBuild an organization ready to experiment, learn, and adapt alongside AI systems.
Manage risk and ethicsLead responsibly by setting guardrails around bias, privacy, and data use.
Drive compounding ROIAI investments grow in value over time as systems learn from new data and integrate across business functions.
In short, AI literacy is becoming as essential as financial literacy — every leader needs to strategically understand how AI changes the economics of work. ### How AI and LLMs Work: A Simple Breakdown **What is a Model?** Think of it like a recipe trained by reading millions of cookbooks. The more data it reads, the better it guesses what’s next — whether it's text, speech, or images. **What is an LLM?** A **Large Language Model** is an AI that has read vast amounts of human text (books, articles, websites) and learned to **predict and generate** language. Example: Ask, “Summarize this contract,” and the LLM scans the text, identifies key points, and writes an easy-to-understand version. **How They “Learn”** AI models don’t “understand” like humans do. Instead, they find patterns and predict what comes next. - **Training**: Feed it tons of data - **Fine-tuning**: Specialize it for specific industries (law, medicine, etc.) - **Inference**: Use it to answer or create something new - **Reasoning** **(new phase):** Modern models are now learning to think through steps, explain their logic, and correct themselves — similar to how humans reason through a problem #### The Reasoning Model Reasoning models go beyond prediction — they plan, verify, and justify their responses. They use structured thought patterns (often called “chains of reasoning”) to: 1. **Break** a complex problem into smaller steps 2. **Evaluate** whether their initial answer makes sense 3. **Generate** explanations and alternatives before responding This is the beginning of AI systems that can explain their thinking, not just produce outputs — a foundational step toward trust and collaboration between humans and machines. ### Business Uses of AI Today AI is already transforming work across industries — not by replacing people, but by taking on the long tail of everyday tasks that span communication, coordination, and decision-making. Leading companies are building AI capabilities into the core of their operations, creating a scalable advantage. | Business Area | Use Cases | Value | Examples | | --- | --- | --- | --- | | Customer Service | Chatbots, auto-replies, email triage, knowledge bases | 24/7 support, reduced workload | Instacart, Vodafone, Amtrak | | Marketing | Copy generation, A/B testing, campaign analysis | Faster execution, tailored messaging | Unilever, Sephora, HubSpot | | HR | Resume screening, onboarding Q&A | Streamlined hiring, better employee experience | SAP, L’Oreal, LinkedIn | | Operations | Document summarization, invoice processing | Time savings, fewer errors | PwC, GE, KPMG | | Sales | CRM updates, proposal drafts, meeting summaries | Stronger engagement, better follow-up | Salesforce, Coca-Cola, Intercom | | Asset Management | Portfolio analysis, client reporting, market research summarization, compliance monitoring | Improved decision-making, faster insights, reduced analyst workload | BlackRock, Goldman Sachs, JP Morgan Asset Management | **Note:** While many companies start with one high-impact use case (like automating support tickets or generating emails), leaders are shifting to AI platforms that serve multiple workflows across departments. For example: - **Morgan Stanley** uses a centralized AI assistant trained on their internal knowledge base, serving wealth managers, compliance teams, and client service. - **Shopify** uses AI to assist both merchants and internal staff across customer support, storefront design, and logistics. This platform approach avoids fragmented tools, boosts consistency, and creates a compounding ROI — each new AI use case adds value to the last. ### Glossary: Demystifying the Jargon
TermUse Cases
AI (Artificial Intelligence)Machines mimicking human intelligence (learning, reasoning, decision-making)
ML (Machine Learning)Subset of AI where algorithms learn patterns from data to make predictions or decisions
LLM (Large Language Model)A model trained on huge volumes of text to generate and understand language
GPTA popular type of LLM developed by OpenAI (e.g., GPT-3, GPT-4, GPT-4o)
TrainingTeaching an AI by feeding it large datasets to learn from
Fine-tuningFurther training of a model for specific tasks, industries, or datasets
InferenceUsing the trained AI model to generate outputs (e.g., answer a question, write an email)
TokenA piece of text (word or part of a word) that the model uses as its basic unitPromptThe instruction or input you give to an AI model to get a responseAPI (Application Programming Interface)A method for connecting software to the AI so businesses can use it in their workflowsAgentAn advanced AI that can take actions, interact with software/tools, and follow multi-step instructionsMultimodalAI that works with multiple input types (text, image, video, audio) togetherRAG (Retrieval-Augmented Generation)Combines LLMs with real-time data from your documents or systems, improving relevance and accuracyCopilotA virtual AI assistant embedded into tools (like Microsoft Copilot or GitHub Copilot) that helps users with tasksMCP (Model Context Protocol)An open standard that lets AI assistants securely connect to external tools and data sources (e.g., files, databases, SaaS apps) via a consistent interface.OrchestrationManaging how multiple AI models, tools, or APIs interact in a coordinated way to complete tasksGuardrailsRules or controls that prevent AI from making harmful, inaccurate, or unauthorized outputsLatencyThe time delay between making a request to the AI and receiving a responseGroundingLinking AI responses to real data or sources to improve accuracy and reliabilitySynthetic DataArtificially generated data used for training models when real data is limited or sensitive
### Final Thought: Where Do You Start? The evolution from AI 1.0 to AI 2.0 highlights one clear lesson: leaders don’t need to build the intelligence — they need to learn how to apply it. You don’t need to build your own AI. You just need to know: - **What** **problem** **you want to solve** - **What** **data** **you have** - **Which** **partner** **will you work with** Start small — test tools with a clear ROI and scale from there. But remember: adopting AI isn’t just a technology decision, it’s a leadership decision. The goal isn’t to chase trends, but to embed intelligence into how your organization thinks, operates, and learns. The companies that win the next decade will be those that treat AI not as a project, but as a core capability — built, measured, and evolved across every function. [![](/assets/blog/an-introduction-to-ai__thirdPartyMemberAvatar-65f078be000ce97ff60d1df9-fe43e052-d280-4a5f-93aa-c7dad2d3b8d8)Guest User](/resources?author=666980ad3edad116e2ada8f6) ## From ETL to Intelligence: Why Clean Data Won’t Be Enough in the Age of AI URL: https://www.travtus.com/resources/from-etl-to-intelligence-why-clean-data-wont-be-enough-in-the-age-of-ai Date: 2025-10-31 Summary: Data pipelines have become the backbone of the modern tech stack, orchestrating and moving data throughout an organization. As AI evolves, so too are these systems. By Andrew Day, AI & Technology Executive on October 29, 2025 Data pipelines have become the backbone of the modern tech stack, orchestrating and moving data throughout an organization. As AI evolves, so too are these systems — shifting from mechanical assembly lines to cognitive threads that think, learn, and connect the enterprise together. I started my career at Experian, the global data powerhouse operating credit bureaux around the world. Back then, data pipelines were built like fortresses — mainframe-based ETL processes (Extract, Transform, Load) designed to move data through tightly controlled systems. Every field was mapped, every rule reviewed, every outcome tested. Precision mattered more than speed, and change was measured in weeks. These pipelines were engineered for reliability, not agility — monuments to a world where control defined confidence. And for that world, it worked. Clean, structured data was the gold standard. ### **From Structure to Understanding** At Travtus, around 2018, the world started to shift. We began experimenting with BERT, one of the first modern language models. At first, it was a small experiment — automating transactional, conversational processes. But the power was clear almost instantly. We could train models to classify, extract entities, and generate metadata. For the first time, we could structure the unstructured — transforming messages, emails, and notes into data that revealed why people were interacting. It wasn’t fast. Labeling data took weeks. Training took months. But it opened the door to understanding. Suddenly, data pipelines weren’t just about moving information — they were orchestrating intelligence. ### **The Generative Leap** Then came Generative AI, and the whole picture accelerated. We could now generate synthetic data, train models faster, and build richer metadata. But it was when large language models developed reasoning capabilities that the real leap occurred. Pipelines evolved beyond tagging or classification. They could summarize, extract, interpret, and even answer questions. And not just for single messages — but across entire customer journeys. Now, a single message could be understood in context — part of a larger story, connected to patterns of behavior and outcomes. ### **Pipelines That Reason** By combining LLMs with retrieval and contextual memory, pipelines have started to reason — understanding not just what data says, but why it matters. We have moved from static data flows to semantic systems — pipelines that can read, comprehend, and relate. They no longer just describe the past. They have begun to understand the present and even predict the future — how people behave, how systems respond, where friction or opportunity lies. Data pipelines are no longer infrastructure. They are systems of reasoning. ### **The Rise of Agentic Pipelines** Now we’re entering the next phase: agentic pipelines. These systems don’t just follow instructions — they make decisions. They can look at a dataset and ask: > “Do I have the information I need?” > “Should I retrieve more context?” > “Do I need to call another model or take an action?” They can reflect, adapt, and re-evaluate. It’s a profound shift — from pipelines that were designed to those that can design themselves. From flows that execute, to agents that think. These are not passive systems anymore. They’re collaborators. ### **The Illusion of Clean Data** I often hear companies say, > “We’ve invested heavily in our data — making sure it’s structured and clean.” And I can’t help but think: You’ve spent a lot of money building pipelines for the world that was, not the world that’s coming. For decades, clean data was the goal because pipelines couldn’t think. Every transformation, every rule, had to be designed by people. But now, pipelines can review data, write their own cleaning logic, pull context, and progress autonomously. The focus is shifting from data cleanliness to data cognition — from “Did we clean this correctly?” to “Did we understand this correctly?” The winners won’t be the companies with the cleanest data, but the ones with the smartest systems. ### **The Mindset Shift** This isn’t just a technological step forward. It’s a mindset shift. For years, we’ve designed systems for control. Now, we must design them for intelligence. The companies that thrive in the next decade will see pipelines not as plumbing, but as living cognitive threads — running through their organizations, constantly reasoning, sensing, and learning. The pipes no longer carry data. They carry understanding. And that changes everything. ## Travtus Partners with Engrain to Bring Spatial Intelligence to Its Everyday(AI)™ Platform URL: https://www.travtus.com/resources/ecosystempartner-engrain Date: 2025-07-01 Summary: Travtus integrates with Engrain, bringing interactive property maps into the Everyday(AI)™ platform. Teams can view operational data on dynamic property maps, adding location context to daily workflows for faster, clearer decisions across their communities. Travtus, the Everyday(AI)™ platform for multifamily operations, today announced a partnership with [Engrain](https://c212.net/c/link/?t=0&l=en&o=4459906-1&h=984358084&u=https%3A%2F%2Fwww.engrain.com%2F&a=Engrain), integrating their interactive map technology into its Gateway interface. This addition introduces interactive property maps to the platform, allowing users to engage with operational data in a visual format. By embedding location context into the daily workflow, Travtus enhances how teams understand, navigate, and act on insights across their communities. Through the Travtus Gateway, teams access customer and community intelligence across their portfolio. One example is Community Profiles, where they can view summaries of resident feedback, complaints, incidents, and escalations. With the addition of maps, teams can now see exactly where issues are occurring. This is especially valuable for regional and centralized teams who are not on-site. Maintenance tasks can also be viewed by location, making it easier to plan and prioritize responses. The maps also support automated leasing conversations by presenting available units to prospective residents. For example, if someone is searching for a three-bedroom unit, the Travtus AI agent can now share real-time unit availability alongside a floor plan. This creates a more informed and engaging experience for prospective renters. By combining insights and automation with property visuals, the platform enables faster, more informed decisions across all functions. This integration supports Travtus' vision of delivering intelligence that is not only real-time but also intuitive and embedded in the way property teams operate. "We are excited to partner with Engrain, one of the first integration opportunities we identified for our Ecosystem," said Tripty Arya, CEO at Travtus. "This collaboration adds a spatial layer to the intelligence we're building into our customers' operations. It reflects our commitment to delivering insight where teams need it most, in a way that is visual, contextual, and part of their every day." "We've always believed maps should be a core part of every platform that powers property operations," said Brent Steiner, CEO of Engrain. "Travtus is helping make that a reality by giving teams one more way to access and benefit from the maps they already use." The integration is available now to all Travtus customers using Engrain's products. Originally published by Engrain on [PR Newswire](https://www.prnewswire.com/news-releases/travtus-partners-with-engrain-to-bring-spatial-intelligence-to-its-everyday-ai-platform-302494767.html). ## Why AI is Essential for Your business URL: https://www.travtus.com/resources/why-ai-is-essential-for-business Date: 2025-01-21 Summary: AI is no longer a future concept—it’s already transforming leasing, resident experiences, and asset management. Whether you're well-versed in AI or just beginning to explore its potential for your business, this article outlines five practical reasons to adopt AI into your 2025 roadmap today.  AI is no longer a future concept—it’s already transforming leasing, resident experiences, and asset management. Whether you're well-versed in AI or just beginning to explore its potential for your business, this article outlines five practical reasons to adopt AI into your 2025 roadmap today.  For an introduction to AI, check out our [AI 101 Guide_._](https://travtus.com/resources/an-introduction-to-ai) ## Five reasons you should embrace AI now Understanding AI's capabilities is a start, but recognizing why it matters and implementing it in your business is essential. ### 1) Access your data Many businesses at the start of their AI journey don’t fully appreciate the value of the unstructured data they currently possess. This data is often siloed across different systems, programs, or teams and isn’t being used to improve products or business processes. Data freedom is incredibly important, for your business and your customers, so unlock the power of the data you already have by incorporating AI.  ### 2) Understand your customers Most of your data is about your customers. While you have this information, it's unlikely you'll sift through it all to figure out how to better serve each prospect or resident. The right AI platform can surface all this information for you. Whether you want to take a personalized approach or better understand how your customers feel, AI delivers tailored and actionable insights seamlessly. Happier customers mean better retention, which in turn improves revenue generation. ### 3) Streamline business operations There are many ways AI can support internal operations and boost your team’s efficiency. Automations can reduce manual work so your team can focus on serving customers. In addition, AI can also streamline business assets and messaging, ensuring a consistent brand voice across all communications. Our speciality lies in unlocking information; businesses gain insights into their customers and we provide informed actions to take, helping property managers understand their priorities and tasks. ### 4) Save time and money Having a more efficient team not only makes the day-to-day running of your business easier, it also saves you time and money. With AI, you can quickly spot and resolve issues, whether through people or automation, which boosts customer satisfaction and reduces turnover. AI also makes it easier to centralize operations, allowing you to uphold service standards while cutting onsite costs. By reducing time spent on unnecessary or manual tasks, your team becomes more productive, allowing your business to grow further and faster. ### 5) Stay competitive Now that you understand AI's impact on your business, the crucial question is: Why prioritize AI now? AI has been a buzzword for years, but we’re entering a critical phase.  Many businesses are starting to adopt AI, so it’s essential to explore your options so you stay ahead of the curve. Seize this opportunity to outpace the competition by implementing true AI that leverages your unique business information for the maximum benefit of your operations. Implementing AI now for the future benefit of your business and customers is why timing is so important. At Travtus, we have positioned ourselves three years ahead of the AI curve by building a platform that utilizes all business data sources, unlike the ‘plug and play’ options that offer limited potential. The potential of AI to transform your business is immense. If you're ready to get started, we’re here to help. ## An Introduction to AI URL: https://www.travtus.com/resources/an-introduction-to-ai Date: 2025-01-17 Summary: AI has become a buzzword across industries over the last many years. While opinions range from enthusiastic support to cautious skepticism, it's important to acknowledge the significant opportunities this technology offers. AI has become a buzzword across industries over the last many years. While opinions range from enthusiastic support to cautious skepticism, it's important to acknowledge the significant opportunities this technology offers. Understanding the basics of AI can unlock valuable insights for businesses looking to innovate and boost operational efficiency. This guide introduces key types of AI and their applications, providing baseline knowledge to help evaluate AI opportunities. ## What is AI? Artificial intelligence (AI) is an umbrella term used to describe the use of machines to mimic human behaviour. You will likely be familiar with the use cases below and may even use them, even if you don’t recognise the AI type. | Type of AI | Definition | Use Cases | | --- | --- | --- | | Natural Language Processing (NPL) | Computers can understand and talk to humans in plain language | • Chatbots and Virtual Assistants e.g. Siri and Alexa
• Text Generation and Summarization e.g. ChatGPT
• Text Analysis
• Language Translation
• Sentiment Analysis
• Speech Recognition e.g. Siri and Alexa | | Computer Vision | Machines can identify and recognize objects | • Image and Video Recognition e.g. medical imaging
• Facial Recognition e.g. smartphone unlocking
• Video Analysis e.g. security and traffic monitoring
• Object Detection e.g. self-driving cars
• Augmented Reality (AR) e.g. gaming
• Optical Character Recognition (OCR) e.g. searchable documents | | Machine Learn and Predictive Analytics | Computers can learn from data and make smart informed forecasts | • Recommendation Systems e.g. Netflix
• Fraud Detection e.g. in banking and insurance
• Diagnostics e.g. in healthcare
• Forecasting e.g. energy consumption
• Risk Assessment e.g. credit scoring systems | | Robotics and Automation | Machines can execute tasks automatically for operational efficiency | • Industrial Automation e.g. robotic arms
• Robotic Process Automation (RPA) e.g. invoice processing
• Autonomous Vehicles e.g. self-driving cars
• Drones e.g. delivery drones
• Collaborative Robots (Cobots)
• Humanoid Robots e.g. customer services bots | | Data Analytics and Business Intelligence | Machines can extract valuable insights from data to make informed business decisions | • Business Intelligence Tools
• Pattern Recognition e.g. customer segmentation
• Analysis e.g. social media monitoring
• Anomaly Detection
• Optimization e.g. supply chain | _Source:_ [_https://www.pwc.be/en/fy25/documents/ai-in-a-vb.pdf_](https://www.pwc.be/en/fy25/documents/ai-in-a-nutshell.pdf) Some of the use cases above are already commonly used, others are still in their nascent stages. You’re unlikely to need to use all of them, but having an understanding of the different examples can help determine the best uses for your business. Now that you know a bit more about AI, if you're curious about its benefits for your business, you can read more in our article [here](https://travtus.com/resources/why-ai-is-essential-for-business). If you want further support on building out your AI roadmap, speak to our experts. [![](/assets/blog/an-introduction-to-ai__thirdPartyMemberAvatar-65f078be000ce97ff60d1df9-fe43e052-d280-4a5f-93aa-c7dad2d3b8d8)Guest User](/resources?author=666980ad3edad116e2ada8f6) ## Harnessing the Power of Conversational AI: A Game-Changer for Multifamily Operations URL: https://www.travtus.com/resources/harnessing-the-power-of-conversational-ai-a-game-changer-for-multifamily-operations Date: 2024-03-26 Summary: Conversational AI brings a spectrum of benefits to improve efficiency in multifamily real estate operations. Let’s explore these benefits in detail. In the dynamic arena of multifamily real estate, the adoption of Conversational AI stands as a beacon of innovation, promising not just to enhance customer interactions but to revolutionize operational frameworks. The essence of tapping into the full potential of conversational AI transcends the basic metrics like “deflection rate.” Given the intricate landscape of multifamily operations, a deeper, more nuanced evaluation of these solutions is imperative for unlocking their comprehensive benefits. ### **The Operational Efficiency Enhancements of Conversational AI** Conversational AI brings a spectrum of benefits that collectively contribute to improving operational efficiency in multifamily real estate operations. Let’s explore these benefits in detail: #### **Broad Scope of Understanding** With the ability to address over 2,000 user needs and 250 processes, a robust conversational AI solution can deeply understand and navigate the complex landscape of multifamily operations. This comprehensive grasp ensures that a wider array of tenant needs are met promptly and efficiently, reducing the strain on specialized human resources. > _To put this in to context, AI solutions from banks usually handle ~300 user need putting_ **_Multifamily support operations 6x more complex as a bank_**_._ #### **Knowledge Mining** By efficiently handling routine inquiries and information requests, conversational AI allows your human resources to concentrate on more complex and valuable activities. This not only hastens response times for common queries but also elevates the job satisfaction of your team by enabling them to focus on high-impact work. Due to the very broad set of topics that customers need information on, efficient question answering requires a comprehensive knowledge base to be created. A spreadsheet of structured data from your CRM is barely going to scratch the surface. You need to find solutions that automate the mining of knowledge and don’t rely on asking hundreds of questions to the onsite teams. #### **Efficiency through Information Capture** When residents request a service or report an issue, conversational AI’s capability to ask targeted questions for information capture drastically reduces the workload on the backend. This accurate data collection at the first point of contact minimizes the need for follow-ups, streamlines task execution, and significantly cuts down the time spent by teams in gathering necessary details. > _With over 250 processes there are few industries that can benefit as much from upfront information capture to ease downstream administration._ #### **Intelligent Auto-Routing** Conversational AI can automatically categorize and route service tickets or inquiries to the most suitable team or individual based on the issue’s nature. This auto-routing ability diminishes the need for a first-level support team, thereby optimizing the operational structure and ensuring that specialized resources are allocated to tasks that genuinely require their expertise. Multifamily is early on its adoption of centralization. If you look to other industries that have adopted centralized models, their constant search for optimized workflow leads to the creation of level 1 support teams that triage and distribute work to specialized teams. This is especially true where you have a hybrid onsite/central structure (The majority of the industry). > _Multi-family operations have the opportunity to step past this and benefit from the_ **_automated L1 support_** _creating more efficient centralized operations from the start._ #### **Future-Proofing Through Potential Automation** While evaluating conversational AI solutions, the scalability through integration with APIs or Robotic Process Automation (RPA) should be a key consideration. These integrations promise to automate more complex tasks in the future, further reducing manual workload and optimizing operational efficiency. > _Even if you do not have the API’s or RPA functions available today, the strategic evaluation of an AI platform should factor this in. If you aspire to increasing automation and self service of customer support, it is critical to be building the right foundations and not narrowing your focus to short term benefit._ #### **Crafting a Future with Conversational AI** For multifamily executives and operations teams, the journey towards integrating conversational AI is one of strategic foresight and comprehensive evaluation. It’s about looking beyond immediate efficiencies to understand how these solutions align with long-term operational goals and technical infrastructure. By embracing a holistic approach that considers both the operational benefits and the profound impact on customer experience, multifamily operators can unlock the true potential of conversational AI. This not only leads to enhanced operational efficiency and customer satisfaction but also lays a robust foundation for innovation and growth in the competitive landscape of multifamily real estate. ## The History of Large Language Models URL: https://www.travtus.com/resources/the-history-of-large-language-models Date: 2023-05-12 Summary: Language Models have been around for a while, but it’s only since ChatGPT’s release that Large Language Models have really captured the public’s imagination. How did Large Language Models come to be, how do they work, and with the emergence of industry-specific language models, what does the future hold? ## What is a Language Model? > “Writing comes from reading, and reading is the finest teacher of how to write." Annie Proulx Just like a human author, a Language Model can construct output based on texts that it has read. Separate from the writing “rules” learned in a classroom, a human can also unconsciously transfer vocabulary, spelling, grammar and stylistic voice learned from a lifetime’s reading to their writing. Likewise, a language model has a pre-read set of written data, from which it can create its own rules. ### The Cat Sat On The… In the AI world, a language model takes examples of writing (a set of data), to give the probability of a subsequent word sequence being “valid.” Don’t get confused by the word “valid”. It doesn’t refer to grammatical or factual validity. Instead, valid refers to the closeness of the sequence to entries in the original dataset. This validity score can be used to generate new content. A simple and obvious example of a language model is predictive text. Consider a language model trained on written data and the incomplete sentence: “The cat sat on the…” A language model might consider a series of alternative endings and match them to patterns in its dataset with various endings being more commonly used than others. The language model doesn’t have any concept of whether a cat might prefer the sidewalk to a mat, or what a cat even is, but - depending on its data - will be able to see which sentence ending is the most probable (or “valid”). Thus, “mat” is predicted before “sidewalk”. ## Large Language Models The first language models trace their roots to the early days of AI. ELIZA, a tool built to converse, debuted in 1966 at MIT. Whilst ELIZA hasn’t learned from data and doesn’t predict the probability of a word or a sentence in language, it’s widely credited with being the first AI to be able to hold a human-like conversation. Many of today’s “chatbots” are built using a similar technique, scripted responses being served to the user according to identifiable words and phrases in the input text. ### Brown Corpus While Eliza relied on canned and deliberately open responses, other AI Language models began to use libraries of written texts to generate output. One of the earliest datasets used by language models was the "[Brown Corpus](https://www1.essex.ac.uk/linguistics/external/clmt/w3c/corpus_ling/content/corpora/list/private/brown/brown.html)," which was created in the early 1960s. The Brown corpus consists of 500 texts, each consisting of just over 2,000 words. Since then, Language models have increased not only in the sophistication of the models used to train the data but also in the size of the data sets themselves. This size increase is primarily due to the availability of large amounts of text data on the internet, coupled with advances in computing power and storage. As an example of the amount of data available, the [Common Crawl](https://commoncrawl.org/) project has been crawling the web since 2008, and currently contains over 70 petabytes of data, much of it in the form of text. To put the magnitude of the Common Crawl project in perspective, some projections suggest that a Petabyte is comparable 500 billion sheets of standard printed text. At the same time, advances in computing power and data storage have made it feasible to train larger and more complex language models. In particular, the development of specialized hardware such as [Google's Tensor Processing Units](https://cloud.google.com/tpu/docs/tpus) (TPUs) has made it possible to train models with billions or even trillions of parameters. While the term “large” is vague, the dataset for training a Large Language Model typically has at least one billion parameters (the variables present in the model on which it was trained.) ### The Birth of GPT The first large language model was the GPT (Generative Pre-trained Transformer) model, developed by OpenAI in 2018. It had 117 million parameters and was trained on a massive amount of text data from the internet. This model was ground-breaking, generating human-like text and performing a range of natural language processing tasks with high accuracy. Since then, larger and more powerful language models have been developed, including GPT-2, 3 and 4. GPT4 is rumored to have 1 trillion parameters. Other big players have joined the scene, with Google, Meta and Amazon all working on their own Large Language Models. ## Large Language Models Timeline
2014Google buy London based AI company, Deepmind 
2018BERT by Google
An influential language model not built to generative text		340 million parameters
 GPT by OpenAI
Ground-breaking LLM that generated human-like text		117 million parameters
2019Microsoft invests $1billion into OpenAI 
2019GPT21.5 billion parameters
2020GPT3175 billion parameters
Dec 2021Gopher by Deepmind300 billion parameters
May 2022ChatGPT released
Built on a fine-tuned variant of GPT3, ChatGPT brings LLMs to the masses.		 
July 2022BLOOM by Hugging Face
Essentially GPT-3 but trained on a multi-lingual corpus		175 billion parameters
Nov 2022Galactica by Meta
Trained on scientific text		120 billion parameters
Nov 2022AlexaTM by Amazon20 billion parameters
 Microsoft invests a further $10 billion into OpenAI 
Mar 2023GPT4 by OpenAI. Also available through ChatGPTRumoured to have 1 trillion parameters
Mar 2023BloombergGPT
LLM trained on financial data from proprietary sources		50 billion parameters
### Things Suddenly Get Real Obviously, the size of the dataset correlates to the accuracy of the language model, and the emergence of Large Language Models (LLMs) with huge datasets has driven a significant increase in performance. LLMs have shown improvements in their ability to generate human-like text, as well as their performance on a variety of natural language processing tasks. But the introduction of large datasets has also seen a "discontinuous phase shift," a stage whereby the model [suddenly acquires abilities](https://upload.wikimedia.org/wikipedia/commons/5/57/LLM_emergent_benchmarks.png) disproportionately larger than the increase in dataset size. Adam, Travtus’ AI teammate, uses LLMs to create conversational intelligence for real estate operations. By combining the embeddings of large models with Travtus’ own property and industry-specific data, Adam is able to understand and respond to resident queries in natural language and with accuracy. But progress has not all been plain sailing. The advances in LLMs have been intercut with embarrassing PR fails for some major companies. Amongst others, Meta’s Galactica, an AI claiming to “summarize academic papers, solve math problems, generate Wiki articles, write scientific code, annotate molecules and proteins, and more.” was released to a chorus of disapproval from its intended userbase, quick to point out its [factual failings](https://www.technologyreview.com/2022/11/18/1063487/meta-large-language-model-ai-only-survived-three-days-gpt-3-science/). ## Learn From the Masters: The Emergence of Domain Specific Large Language Models > “Find an author you admire and copy their plots and characters in order to tell your own story…” Michael Moorcock As humans master writing by focusing on the works they aspire to emulate, so the recent development of Domain Specific LLMs have sought to improve real-world performance with more specialized datasets. The complexity and unique terminology of the domains such as Real Estate and Finance warrant a domain-specific model. While the GPT family of LLMs were pre-trained over a huge amount of text from many different sources such as Common Crawl, webtexts, books, and Wikipedia, domain-specific Large Language Models generally use a separate proprietary dataset combined with a pre-existing LLM. BloombergGPT is one such example ### BloombergGPT Purpose built for finance natural language tasks, [BloombergGPT](https://www.bloomberg.com/company/press/bloomberggpt-50-billion-parameter-llm-tuned-finance/) has access to the vast amount of data available on the Bloomberg Terminal, allowing it to outperform similarly-sized open models on financial natural language tasks while still being able to match performance on non-specific natural language tasks. Where Bloomberg have led, other industries are following, curating their own collections of data to create LLMs that can generate text that is not only human-like but accurately represents their own domain-specific knowledge. ### Other Domain Specific Large Language Models
BioMedLM by Mosaic ML and StandfordA purpose-built AI model trained to interpret biomedical language.
BloombergGPT by OpenAI and BloombergPurpose built for finance-based natural language tasks
Galactica by MetaBuilt for the scientific community
CodexFine-tuned on publicly-available Python code
**The Future of Large Language Models** To the general public perhaps, it may seem like the technology behind Large Language Models is only a few years old. Actually, language models have been around since the 1960’s and the idea of creating writing based on a dataset of existing text has existed for as long as human authors have had access to the written word, created 5500 years ago in Mesopotamia.) > **“**There is no such thing as a new idea. It is impossible. We simply take a lot of old ideas and put them into a sort of mental kaleidoscope. We give them a turn and they make new and curious combinations. We keep on turning and making new combinations indefinitely; but they are the same old pieces of colored glass that have been in use through all the ages”. Mark Twain However, the creation of Large Language Models and the freely available release of ChatGPT means that Large Language Models have taken center stage in the public consciousness. No surprise, as larger sets of data filtered against bias, domain-specific LLMs , and advances in computing power and data storage promise to further improve human efficiency and productivity. The future of LLM’s is bright. ## Quick Read: What AI Means For Your Community Managers URL: https://www.travtus.com/resources/quick-guide-what-ai-means-for-your-community-managers Date: 2023-05-03 Summary: AI can help real estate owners and operators manage their properties more efficiently, but what does this AI revolution mean for onsite staff and community managers? The use of artificial intelligence in property management is becoming increasingly popular. AI can help real estate owners and operators manage their properties more efficiently, but what does this AI revolution mean for onsite staff and community managers? ## Increase in Staff Retention and Specialization One benefit of AI tools like Adam is an increase in efficiency. AI can help community managers save time and money by automating tasks. Advanced AI tools like Adam can answer, respond to, and action resident inquiries. This frees up time for community managers to specialize in their roles - concentrating on the needs of their residents and team, increasing job satisfaction and staff retention at the same time. ## Decisions Will Be Backed By Data AI can also use data-driven insights to help property managers better understand their residents, identify potential issues, and make more informed decisions, based on data, rather than supposition. Community managers can also use this data to identify and communicate pain points to higher management. The long tail of unique resident inquiries that takes up so much time for community managers is now a tangible target. Using Adam, these unique resident problems can be tracked, communicated and solved. ## Improved Resident Retention and Community Experience Improvement in conversational intelligence means that frustrating chatbots are on the way out, to be replaced by resident-friendly conversations through multiple channels. AI tools like Adam don’t need hours of scripting either, instead bringing with them years of industry best practice and location-specific knowledge. AI can be used to automate marketing and outreach efforts, allowing property managers to focus on building relationships with residents and improving resident retention rates. ## Travtus: Real Estate and Technology Investor’s Pick URL: https://www.travtus.com/resources/travtus-real-estate-and-technology-investors-pick Date: 2023-03-28 Summary: Travtus is featured in Business Insider's listing of the most exciting Proptech startups. The list is decided by more than twenty real-estate and construction technology investors and we’re very proud to be there. Thank you to everyone involved. Travtus is featured in [Business Insider's listing of the most exciting proptech startups](https://www.businessinsider.com/top-proptech-companies-startups-2023-3). The list is decided by more than twenty real-estate and construction technology investors and we’re very proud to be featured. Thank you to everyone involved. ## Turbocharge ChatGPT for Your Community Managers URL: https://www.travtus.com/resources/turbocharge-chatgpt-for-your-community-managers Date: 2023-03-24 Summary: When ChatGPT is integrated into a system designed specifically for Community Managers, ChatGPT can become a small, highly functioning part of a much more powerful tool. ## What is ChatGPT? ChatGPT is an artificial intelligence (AI) chatbot developed using OpenAI’s GPT3 language model. Earlier versions, pre-trained with general data from the internet and social media, have been fine-tuned, allowing it to generate increasingly natural responses to user input. ## How Can Community Managers Use ChatGPT? Through its publicly available interface - ChatGPT is a powerful language tool for everyone. Community managers can use it to produce written content quickly if they provide the correct prompt. As an example, we asked it to write marketing copy with the following prompt: _“Write a paragraph describing a luxury two-bedroom apartment in a busy part of town. Its selling points are nearby cafe culture, vibrant nightlife and off-street parking.”_ And this is what it came up with… _“This stunning two-bedroom apartment is situated in the heart of a vibrant and exciting part of town. With its close proximity to popular cafes, bars, and restaurants, you’ll never be short of things to do. Whether you’re looking for a night out or just want to relax in one of the many nearby coffee shops, you’ll be spoiled for choice. It’s also perfect for those who need off-street parking, as there is plenty of secure parking onsite. With its chic interior design, this luxury apartment will make you feel right at home. Enjoy the convenience of being close to the action, with all the amenities you need for a comfortable life”._ We prompted it again, this time with a more awkward problem… “_Write a quick note to an angry resident apologizing for continued power outages.”_ It returned the following… “_Dear \[Name\],_ _We understand your frustration with the ongoing power outages in your area. We apologize for any inconvenience this has caused you and your family._ _Our team is actively working to resolve the power outages as quickly as possible. We appreciate your patience and understanding._ _Sincerely, \[Your Name\]”_ Once again, not perfect, but certainly not bad. ## The Limitations of ChatGPT The limitations of ChatGPT are clear, though. Take this prompt, for example. _“Please list the rental costs of apartments in \[a real address was supplied\] for a prospective resident.”_ It returned… _“The rental costs for \[the address\] are as follows:_ _Studio: $2,495_ _1 Bedroom: $2,750_ _2 Bedroom: $3,350_ _3 Bedroom: $4,450”_ A quick google search revealed that these rental costs were incorrect. ChatGPT had filled in the blanks of its knowledge in the only way it knows, bluffing. With that in mind, look back at the first example in the previous section…. “_With its chic interior design…”_ We didn’t actually specify that the interior design was _chic._ It’s probably just about safe to assume a luxury apartment will have chic interior design, but what if it was funky or industrial? We’d be attracting the wrong type of prospect. We didn’t supply enough information, but rather than leave details out, ChatGPT did it’s best to create a natural response. In short, it bluffed. Now the second reply… “_Our team is actively working to resolve the power outages as quickly as possible.”_ Is it? We didn’t actually say. Maybe the work is scheduled in the future? Perhaps we’re using a third-party team? Maybe we’re not even in a team. Maybe the power outage is permanent and we’re pursuing another course of action. Potentially, ChatGPT has just lied to the resident about the course of action we are taking. If ChatGPT doesn’t have the correct information in its dataset or the prompt is missing important details, it will attempt to bluff the response using whatever information it has available. This information isn’t always correct, and Chat-GPT2 won’t warn you. Unlike a Google search, you won’t be able to verify its sources. It’s easy to see how, in the wrong hands - say, someone asking it to write legal text - this could be very dangerous. In summary, ChatGPT isn’t a communication silver bullet. It still needs someone or something to review and modify its output. ## How to Turbocharge ChatGPT for Community Managers The true power of ChatGPT can be activated when integrated into systems built specifically for Community Management. Adam, for example, has access to extra data - based on conversations between residents and community managers, preloaded industry standards, and other sources that can return accurate information for specific locations. These hard facts come from Adam’s own data sets, while the chat is being generated by Adam’s AI using ChatGPT as a tool to formulate the written content. This is why a resident or prospect can ask Adam a specific question, such as rental costs, and Adam will return the correct information. Add this to Adam’s ability to integrate into existing systems to book maintenance visits, schedule home viewings and much more, and we start to see how powerful it can be for community managers compared to ChatGPT alone. ChatGPT is a valuable tool for generating written content. It can save Community Managers time, especially if writing isn’t a strong point. However, its use as an independent tool is limited. Discretion is needed regarding when to use it and what for. However, when ChatGPT is integrated into a system like Adam, designed specifically for Community Managers, ChatGPT can become a small, highly functioning part of a much more powerful tool. ## Quick Read: GPT4 for Real Estate URL: https://www.travtus.com/resources/quick-read-gpt4-for-real-estate Date: 2023-03-20 Summary: What does the release of OpenAI's GPT4 mean for Real Estate: A quick read. While the rest of the world is buzzing with the capabilities of ChatGPT, the AI world is focused on the introduction of the next iteration of the technology behind it, OpenAI's GPT4. GPT3.5 is the technology behind ChatGPT. OpenAI has recently released GPT4, which, while it doesn't have a free, user-friendly web app like ChatGPT, will further push the capabilities of AI in real estate. OpenAI sums up the advantages of GPT4 over GPT3.5 as solving problems with greater accuracy, thanks to its broader general knowledge and problem-solving abilities. As such, GPT-4 surpasses ChatGPT in advanced reasoning capabilities, outperforms ChatGPT by scoring in higher approximate percentiles among test-takers, is less likely to respond to requests for disallowed content, and is more likely to produce factual responses than GPT-3.5 So what does GPT4 mean for real estate? It's safe to assume that AI tools using GPT4 will continue progressing in conversational ability. More excitingly, Real Estate tools that combine proprietary real estate and location-specific datasets with GPT4 will be able to leverage GPT4's conversational capabilities to understand residents and onsite teams even more accurately. This combination of GPT4 with real estate AI allows these tools (like Travtus’ AI Community Manager, Adam) - not just to provide written responses to human requests and questions - but also to listen, respond and act with even greater accuracy. ## How SVB’s Collapse Effects Proptech URL: https://www.travtus.com/resources/how-svbs-collapse-effects-proptech Date: 2023-03-13 Summary: On Friday, 10th March, Silicon Valley Bank (SVB) failed due to a bank run and capital crisis. It is the most significant US bank failure since Washington Mutual in 2008, leaving investors, customers, and some proptech companies’ futures in doubt. On Friday, 10th March, Silicon Valley Bank (SVB), the go-to bank for US tech startups, suddenly failed due to a bank run and capital crisis. It is the most significant US bank failure since Washington Mutual in 2008, leaving investors, customers, and some proptech companies’ futures in doubt. This article summarizes the bank’s downfall and what could be on the horizon for proptech in general. ## SVB and Proptech It’s common knowledge that [Silicon Valley Bank went big](https://www.svb.com/blogs/dax-williamson/the-228-trillion-opportunity-proptech) with proptech’s "$228 trillion opportunity," saying, "…in real estate, as in other industries, no matter how ingrained and protected the status quo, a forced evolution has proven imminent.” ## What Happened to SVB? The Federal Reserve raising interest rates, the downturn of tech stocks, the decrease in value of long-term bonds, and a lack of venture capital led to the downfall of the banker. The bank had amassed a $21 billion bond portfolio, yielding an average of 1.79%, and as the 10-year Treasury yield rose to 3.9%, their losses began to mount. Withdrawals from the bank also began to accelerate, with venture capital firms advising companies to withdraw their money. The bank’s stock began to plummet and other bank shares were dragged down with it, leading to the bank being shut down and put into receivership by California regulators. TechCrunch’s [take](https://techcrunch.com/2023/03/12/fintech-interchange-svb-implosion-impact/) the following Monday summed up industry fears: “The impact of this event will be severe, widespread and — for fear of being dramatic — potentially catastrophic for many. Already, businesses are worried about making payroll, which could lead to unanticipated closures and layoffs”. ## What’s Happening Now? While the collapse of SVB initially caused panic on Wall Street, analysts have said it is unlikely to cause a domino effect like the 2008 financial crisis. Moody’s chief economist Mark Zandi stated that the banking system is "as well-capitalised and liquid as it has ever been." Meanwhile, US treasury secretary Janet Yellen is working with regulators to respond to the crisis to protect depositors, but isn't considering a major bailout . [She told CBS News](https://finance.yahoo.com/news/1-yellen-working-address-svb-133249079.html?guccounter=1): "Let me be clear that during the financial crisis, there were investors and owners of systemic large banks that were bailed out...and the reforms that have been put in place means we are not going to do that again." ## How Does the Collapse Affect Companies With Exposure to SVB? It’s never good to have exposure to a collapsing bank, even if you manage to extract yourself in time. In a [LinkedIn](https://www.linkedin.com/pulse/impact-venture-debt-svb-collapse-michael-mandel%3FtrackingId=JgKSjLXwvs5vfAGCY1cn6w%253D%253D/?trackingId=JgKSjLXwvs5vfAGCY1cn6w%3D%3D) post, Michael Mandel, the Co-Founder & CEO of CompStak, raised the issue of Venture Debt, the reason many SVB customers couldn’t have moved their money from SBV even if they’d wanted. SVB describes its [venture debt loans](https://www.svb.com/startup-insights/venture-debt/what-is-venture-debt) thus: “Venture debt is a catch-all term referring to loans that are tailored to the needs and the risks associated with investor-backed startup companies in technology, life science and the innovation economy. These loans are targeted toward companies that have raised equity from venture capital firms or similar institutional sources.” But, and it’s a big but considering the collapse, there is a key provision that many Venture Debt facilities have: The company must have its money with the lending bank. Mandel writes “it is a key reason that SVB tripled its deposits in 2021-2022. It's also the reason that many companies did not jump ship on Thursday and Friday. If they moved their money from SVB, they would trip a covenant of their Venture Debt, and they would lose access to the money”. Considering broader market conditions and interest rates, it’s unlikely many companies could replace their venture Debt facility with equivalent terms. In the words of Mandel, “It is the fear of losing access to this debt facility that caused many to choose not to pull their money.” So what does this mean? Simply, many startups' unused Venture Debt facilities are probably wiped out, and it doesn’t take an expert to see how that will cause problems for the companies exposed. On an industry level, the Venture Debt market is likely to change, and probably not in favor of companies seeking finance. ## How Does SVB’s Collapse Affect Proptech? The stabilizing of the banking industry means that proptech firms, without exposure or deposits in SVB, won’t be affected by its collapse. [Travtus founder and CEO Tripty Arya wrote](https://www.linkedin.com/in/triptyarya/), “Amidst the recent developments at Silicon Valley Bank, I want to reassure our customers and partners that Travtus has no exposure or deposits with SVB, and our company cash remains secure. Our operations remain unchanged, and we are committed to providing the same quality of service,” before echoing the thoughts of everybody in the tech industry. “We stand in solidarity with those affected by this situation.” ## The Story Behind the Travtus Logo URL: https://www.travtus.com/resources/the-story-behind-the-logo Date: 2023-02-22 Summary: Travtus is all about taking something complex and making it simple, so when we asked UK-based design agency, Thoughtcraft, to reimagine our logo, this was the problem that their branding expert, George Hadley, would have to address: How do you instill a diverse range of services, values, and an AI teammate who can do everything, into a simple and elegant brand? Travtus is all about taking something complex and making it simple, so when we asked UK-based design agency, Thoughtcraft, to reimagine our logo, this was the problem that their branding expert, George Hadley, would have to address: How do you instill a diverse range of services, values, and an AI teammate who can do everything, into a simple and elegant brand? The result? He knocked it out of the park. After the process, we caught up with George and asked him some quick-fire questions about his process. #### What’s the story behind our new Travtus logo? I researched the story of Travtus with its roots in architecture, construction, finance and technology. I also asked questions about your target market and aspirations. I then created several strategic directions to help inform the new visual identity. The team selected a route that focused on the company’s aim to add value and make positive improvements to systems and ultimately, to people’s lives. The symbol combines a plus, the letter ’T’, and a simplified human figure. It represents the digital workforce, both for real estate and the future. #### What was the best thing about working with Travtus? The turnaround time for the identity was short, but because the team was very quick and thorough in supplying information and responses, the process worked smoothly. The tech world often works in this intense rapid prototyping style, and this worked well to push through the design process at speed. #### Where do you work? I have a studio space in a co-working building in Lewes. A historic English town, set in the hills a stone’s throw inland from the city of Brighton. There are lots of other creative businesses and individuals in the building and a friendly buzzy cafe downstairs when I need some stimulation in the form of caffeine or conversation. #### What’s your daily routine? I’m very lucky to live a 5-minute walk from work, so I have a very easy commute. The working day is a combination of making lists (hopefully crossing some things off), emailing and speaking with clients, scribbling in a sketchbook and tinkering on the computer. I make sure I get some fresh air and a break from the screen at lunch and always make it back to eat with my family in the evening and chat about what we’ve all been up to. #### What brands are inspiring you, right now? Courier - https://www.couriermedia.com Wide-ranging insights about business and culture - traditional magazine plus newsletters and podcasts. Headspace - [https://www.headspace.com](https://www.headspace.com) Bringing mindfulness to the mainstream with a friendly but calm illustrated aesthetic. Art of Ping Pong - [https://www.theartofpingpong.co.uk](https://www.theartofpingpong.co.uk) Fusing creativity and fun while supporting good causes. Do Lectures - [https://thedolectures.com](https://thedolectures.com) Inspiring lectures, events and publications. Finisterre - [https://finisterre.com](https://finisterre.com) B Corp clothing company that makes ethical long-lasting high-quality products. #### What’s next for Thoughtcraft? A variety of interesting projects with a mix of strategy, brand identity, animation, and some more traditional print is in the mix. After a busy few months, I’m also hoping to take some time to work on my own positioning and marketing. Visit Thoughtcraft’s website - [https://thought-craft.co.uk](https://thought-craft.co.uk) See George’s online scrapbook on Instagram - [https://www.instagram.com/thought\_craft/](https://www.instagram.com/thought_craft/) Connect on LinkedIn - [https://www.linkedin.com/in/thoughtcraft/](https://www.linkedin.com/in/thoughtcraft/) ## Gartner Trends: Adaptive AI for Real Estate URL: https://www.travtus.com/resources/gartner-trends-adaptive-ai-for-real-estate Date: 2023-02-01 Summary: Among Gartner’s tech trends 2023 is adaptive AI, an evolution of artificial intelligence that works by learning continuously from new data, making it considerably more agile in the face of real-world change. It’s remarkable and will have an impact on the real estate industry. Among Gartner’s Top Strategic Technology Trends 2023 is adaptive AI, an evolution of artificial intelligence that works by learning continuously from new data, making it more agile in the face of real-world change. ## What is Adaptive AI? Traditionally, all software – including AI – has relied on developers to revise the code and algorithms. Adaptive AI “can revise \[…\] to adjust for real-world changes that weren’t known or foreseen when the code was first written” ([Gartner](https://www.gartner.com/en/articles/why-adaptive-ai-should-matter-to-your-business)). Businesses are increasingly needing to be able to adapt to changing conditions – something highlighted by the extremes of the pandemic, the climate crisis and the destabilizing effects of war. So businesses need adaptable technology. Erick Brethenoux, a distinguished VP analyst at Gartner, says, “Adaptive AI systems aim to continuously retrain models or apply other mechanisms to adapt and learn within runtime and development environments — making them more adaptive and resilient to change.” ## Why is it Important for Business? In today’s world, we are collecting ever more data, from the web, from the internet of things, from past machine experience and past human experience. Adaptive AI can autonomously sift through this data, learn from it, and optimize its application. Adaptive AI decision-making is broader and more flexible, resulting in improved UX and productivity. These are vital in many industries, particularly in real estate and property management, with their ever-changing factors, both human (changing residents) and market (unit availability, market price trends). Digital entrepreneur and founder of [Spread Great Ideas](https://spreadgreatideas.org/), Brian David Crane [explains](https://www.cmswire.com/customer-experience/adaptive-ai-will-improve-the-customer-experience/) adaptive AI: “By analyzing social, behavioral and past interactions, adaptive AI uses continuous interactions to predict and anticipate customer behavior and provide highly personalized solutions.” This looking at the data of human interactions, of consumer behavior – be it purchases online, in a mall, or even choosing a new apartment – will give businesses advantages in real-time, with the adaptive AI learning and adapting continuously, adjusting goals and outcomes. ## Case Study: Dow Inc According to Crane, brands like Amazon, Netflix and Google are already using adaptive AI. But so are other businesses with a less visible presence in the home. Dow Chemicals Company may have been founded in 1897, but it is at the cutting edge of deploying new technologies. This was recognized when it was among the organizations garlanded in the [Business Intelligence Group’s Artificial Intelligence Award in March 2022](https://www.bintelligence.com/blog/2022/3/22/6-people-26-companies-and-65-products-awarded-for-excellence-in-artificial-intelligence). The multinational, which has its headquarters in Michigan, produces chemicals, agricultural products and plastics. The latter’s division, Dow Polyurethanes, deployed its adaptive AI Predictive Intelligence to address customer input directly. It can model product properties in real-time, and Gartner trends report it “has resulted in a 320% increase in value generated by the analytics platform.” ## What’s Next for Real Estate? While business leaders obviously can’t argue with a statistic like that, Adaptive AI will also simply speed up processes. Adaptive AI will therefore be another area of IT that will require investment, as it can’t simply be bolted onto existing AI. Gartner says, “it is critical to develop less rigid AI engineering pipelines or build AI models that can self-adapt in production.” Gartner predicts that businesses adopting adaptive AI will have a performance advantage of at least 25% over their peers. That’s another statistic it’s hard to argue with. Businesses in the real estate and property management market could begin to embrace adaptive AI with the ambition that its real-world flexibility and digital resilience will likely pay higher dividends in industries - like real estate - that are prone to disruption. ## Gartner Trends for Real Estate URL: https://www.travtus.com/resources/gartner-trends-overview Date: 2023-02-01 Summary: Every year, Gartner gives its predictions for trends in technology. These are closely followed by CTOs, to help their companies stay competitive. We asked Travtus co-founder and CTO Andrew Day to pick the Gartner trends he felt were most applicable to the real estate industry. ### Every year, Gartner gives its predictions for trends in technology. These are closely followed by CTOs, to help their companies stay competitive. We asked Travtus co-founder and CTO [Andrew Day](https://www.linkedin.com/in/andrewrobertday/) to pick the Gartner trends he felt were most applicable to the real estate industry. Gartner, founded in 1979, is one of the world’s preeminent technology research and consultation businesses. From its base in Stamford, Connecticut, it works in more than a hundred countries worldwide, closely tracking changes in technology and advising businesses on how to deploy it. Gartner announced its [Top Strategic Technology Trends 2023](https://www.gartner.com/en/articles/gartner-top-10-strategic-technology-trends-for-2023) in October 2022, giving CTOs and business leaders invaluable insight. David Groombridge, distinguished VP analyst at Gartner, says in his introduction to the report that "...Technology is key, but you have to know when and where technology trends will potentially have an impact." From Gartner’s ten 2023 Trends, Andrew chose industry cloud platforms, platform engineering and adaptive AI as the most relevant to real estate and property management. ## Industry Cloud Platforms The Industry Cloud focuses the concept of shared files, data and services stored on remote servers, on specific industries, creating a platform tailored to the needs of an individual company. This speeds up processes, increases agility, improves performance and bolsters resilience to change. > "The industry cloud is the next evolution of cloud computing. Flexible and faster industry cloud solutions, like Travtus' AI teammate, Adam, are emerging to meet the real estate industry's ever-changing needs, faster and more efficiently than ever before." Andrew Day, Travtus Read our longer article on industry [cloud platforms](https://www.travtus.com/explore/the-industry-cloud-for-real-estate). Find out about [Adam](/product). ## Platform Engineering Platform Engineering is all about the link between the company infrastructure’s complex guts and the external user. With the correct company platform in place, developers can address end users' specific needs, curating suitable, integrated packages of tools and processes. This will replace any patchwork of systems a business previously used with bespoke solutions. > "Many real estate owners are outgrowing their current patchwork of off-the-shelf technology. The good news is that cutting-edge real estate businesses are adopting a platform engineering approach. In the next few years, this investment in tech will start to return, company-wide" [Andrew Day](https://www.linkedin.com/in/andrewrobertday/), Travtus. Read our longer article on [platform engineering](https://www.travtus.com/explore/gartner-trends-platform-engineering-and-real-estate). ## Adaptive AI Adaptive AI is an evolution of artificial intelligence. A step where the AI learns continuously from new data, making it more agile and resilient in the face of real-world change. One key element of this improvement to AI is its ability to self-adapt: to "adjust for real-world changes that weren't known or foreseen…" (Gartner). It will also sift ever-increasing amounts of data with more precision to know what is most relevant to its goals. Gartner predicts that businesses adopting adaptive AI will have a performance advantage of at least 25% over their peers. > "If adaptive AI is correctly applied, owners and operators could find a solution that changes with the problem." [Andrew Day](https://www.linkedin.com/in/andrewrobertday/), Travtus Read our longer article on [adaptive AI.](https://www.travtus.com/explore/gartner-trends-adaptive-ai-for-real-estate) ## Gartner Trends: Platform Engineering for Real Estate URL: https://www.travtus.com/resources/gartner-trends-platform-engineering-and-real-estate Date: 2023-02-01 Summary: Among the predictions of Gartner’s Top Strategic Technology Trends 2023 is the increased importance of platform engineering. This approach to an overarching integration of the developer and end-user experiences will be invaluable to the real estate and property management markets. Among the predictions of Gartner’s Top Strategic Technology Trends 2023 is the increased importance of platform engineering. This approach to an overarching integration of the developer and end-user experiences will be invaluable to the real estate and property management markets. > “Many real estate owners are outgrowing their current technology, but the good news is that cutting-edge real estate businesses are adopting a platform engineering approach, and in the next few years, this investment in tech will start to return, company-wide.” [Andrew Day](https://www.linkedin.com/in/andrewrobertday/), Travtus. ## What is Platform Engineering? A general movement in technology trends is towards systems featuring flexibility to personalize individual business needs combined with standardization to improve efficiency. In the case of platform engineering, the main gist is, in the [words of Gartner itself](https://www.gartner.com/en/articles/what-is-platform-engineering), improvements of “the developer experience and productivity by providing self-service capabilities with automated infrastructure operations.” In other words, platform engineering will streamline how software developers work, enabling them to produce business-specific IT solutions more quickly, ultimately optimizing operations for the user, whether that user is a community manager, website contributor or potential client. It’s all about the link between the company infrastructure’s complex guts and the external user. ## How Does it Work? Platform engineering works by optimizing the efficiency of the developer or development team. It is all about integration and interface. Gartner believes 80% of software engineering organizations will create platform teams by 2026. These will provide the services to deliver the applications, ultimately solving “the central problem of cooperation between software developers and operators.” Modern software architecture is increasingly complicated, and businesses like real estate and property management will have built up a patchwork of off-the-shelf and partially customized applications. While this may have offered a short-term solution, it is not an efficient, optimal long-term solution. With platform engineering, developers will address end users’ specific needs, then develop and curate a suitable, integrated package of tools and processes. This will increase productivity and efficiency for end users and reduce their “cognitive burden“ – the stress on human brain power elicited by relying on that older patchwork. ## Case Study: Nike Developers will start with internal developer portals (IDPs) and then extend from there, embedding security in all platforms early in the development process. Viable commercially available engineering platforms will become available with flexible customization. The Gartner report gives an example of the US brand Nike. As a multinational dealing in multiple areas – notably athletic shoes but also other sportswear, equipment and services – the company’s needs are complex and disparate. Gartner reports they are using “composable platforms.” Composable is a tech buzzword referring to a standardized but modular approach to systems development. This structure integrates worldwide elements of the business, resulting in improved scalability, more agility in the marketplace, and reduced operational costs. ## Platform Engineering for Real Estate Ultimately, platform engineering is another deployment of emerging technology that will improve efficiency in complicated businesses - like those of Real Estate Owners and Operators. Simply put, the speed and ease of integrating new systems and the removal of the dependence on vendors to implement change, means end-users from Community Managers to Accountants will have better software at their disposal, optimized for their roles. Paul Delory, VP Analyst at Gartner, [says](https://www.gartner.com/en/articles/what-is-platform-engineering) platform engineering will “help end users and reduce friction for the valuable work they do” thanks to “operating platforms that sit between the end user and the backing services on which they rely.” ## Gartner Trends: The Industry Cloud for Real Estate URL: https://www.travtus.com/resources/the-industry-cloud-for-real-estate Date: 2023-02-01 Summary: Gartner’s Top Strategic Technology Trends 2023 includes industry cloud platforms, a next level of cloud computing explicitly tailored to a business’s needs, which will be invaluable in real estate. Gartner’s Top Strategic Technology Trends 2023 includes industry cloud platforms, a next level of cloud computing that can be tailored specifically to a business’s needs. It’s a trend that Travtus CEO, Andrew Day has picked to become especially relevant to the real estate industry this year. > “The industry cloud is the next evolution of cloud computing. Flexible and faster industry cloud solutions like Travtus’ AI teammate, Adam, are emerging to meet the real estate industry’s ever-changing needs, faster and more efficiently than ever before.” [Andrew Day](https://www.linkedin.com/in/andrewrobertday/), Travtus. ## The History of Cloud Computing The concept of cloud computing has existed for years, becoming a widespread reality towards the end of the aughts, notably with the release and widespread adoption of cloud storage, file hosting and synchronization services like Dropbox and OneDrive. Rather than using data storage on the user’s computer or the company’s servers, cloud computing migrated vast quantities to server centers. The ease of data storage and file sharing transformed business practices. Real estate, like many businesses, benefitted from the changes. Now, a decade and a half after these services emerged, a new form of cloud computing is occurring in the form of industry cloud. According to [Gartner](https://www.gartner.com/en/articles/what-are-industry-cloud-platforms), in 2021, “enterprises were using industry cloud platforms to accelerate less than 10 percent of their critical business initiatives.” Gartner “predicts that by 2027 it will be more than 50 percent.” ## What is the Industry Cloud? The industry cloud is the transformation of generic cloud computing to business-specific cloud platforms hosting an integrated ecosystem of software, apps, systems and data gathering, management and analysis. In the words of the Gartner report, “Industry cloud platforms combine software, platform and infrastructure as a service (IaaS) with tailored, industry-specific functionality that can more easily adapt to the relentless stream of disruptions in their industry.” A key term there is “tailored.” Industries like real estate have particular needs: managing properties, communities, leases and costs in an ever-changing market. With the industry cloud, your IT can be specifically tailored for property management needs. It is more adaptable and versatile, holistic but modular. The word used is “composable” in that you can compose or customize a refined, efficient system from diverse elements. ## Case Study: Hangzhou City Brain The Gartner report gives one fascinating case study of an industry cloud: the Hangzhou Smart City Brain. Hangzhou is a significant Chinese population center (of around 10 million people), long plagued by traffic congestion and inefficient urban systems. The city authorities partnered with Hangzhou-based tech company Alibaba to design the City Brain cloud system, which had its city-wide roll-out in 2017. Utilizing data from traffic signals, GPS, the transportation bureau, the public transportation system and more, it was able to reduce journey times, diminish traffic jams, increase response times to accidents, decrease illegal parking and more. Essentially, it gathered data, processed it algorithmically, then fed it back into management systems (such as traffic signals) around the city, resulting in increased urban efficiency. ## The Industry Cloud for Real Estate The industry cloud is a clear evolution of cloud computing, providing tailored, industry-specific solutions, speeding up processes, increasing agility, improving performance and bolstering resilience to change and uncertainty. In the case of real estate, specialized systems and tools are emerging to meet the specific needs of owners and operators, allowing them to move away from more generic products - traditionally slower to adapt to the changing demands of the real estate industry. ## Using AI to Boost Apartment Operations and Financial Performance URL: https://www.travtus.com/resources/using-ai-to-boost-apartment-operations-and-financial-performance Date: 2023-01-16 Summary: Travtus CEO Tripty Arya talks to Mike Consol about our “Everything Teammate” Adam, and how Adam’s industry knowledge can boost apartment operations and financial performance. Travtus CEO [Tripty Arya](https://www.linkedin.com/in/ACoAAAPmx34BlFkxxSLDqXYUHIPrt88-6vOjZn0) talks to [Mike Consol](https://www.linkedin.com/in/ACoAAAAcApABpnsFUV-9wUNip47vbCchrF9N7_k) about our “Everything Teammate” Adam, and how Adam’s industry knowledge can boost apartment operations and financial performance. ## Beyond The Hype of ChatGPT: How to Unlock AI for Real Estate URL: https://www.travtus.com/resources/beyond-the-hype-of-chatgpt-how-to-unlock-ai-for-real-estate Date: 2022-12-08 Summary: It's hard to ignore the buzz around OpenAI's newest release, ChatGPT, the latest development in the GPT series of super large language models. It’s being hyped by some as a huge labor-saving tool. But how can this technology be leveraged to benefit real estate owners and operators? It's hard to ignore the buzz around OpenAI's newest release, ChatGPT, the latest development in the GPT series of super large language models. It’s being hyped by some as a huge labor-saving tool. But how can this technology be leveraged to benefit real estate owners and operators? The answer lies in AI and real estate teams working together to develop domain-specific solutions, combining the most recent advancements in technology with industry knowledge to create practical solutions for your real estate business. ## What is a Large Language Model? Large language models are a type of machine learning model trained to understand the meaning and context of words and sentences from massive amounts of text data scraped from the internet. You already interact with language models every day: they're used to rank search results, automate translations, and summarize news articles. The auto-complete feature on your phone? That's a language model. Since their development started to accelerate in 2018, the use of language models has been growing at an astonishing rate. A new space race has been ignited, with larger and larger language models being trained. Supercomputers are processing datasets containing billions of words, and companies like OpenAI, Google, Microsoft, Facebook, and Salesforce are investing heavily to stay ahead of the competition. ## What Does it Do? In short, these models are mastering language and memorizing knowledge contained in the training data. The data is so vast that you can ask general knowledge questions and receive coherent answers. The conversational flow is very smooth and feels convincing. It's easy to see why people are starting to wonder if this kind of technology will replace human staff: If AI can create a story, answer general knowledge questions, and write an apartment listing, is helping residents and prospects no problem too? ## The Problems If you look past the hype, even ignoring some well-documented shortcomings of super large language models, including bias (racism, sexism, etc.), overconfidence (or lying), and unpredictability, why aren’t large language models replacing people in real estate? The answer lies in what your team can do that ChatGPT can’t. - Your team has industry knowledge learned, not through reading, but by doing. - Your team has experience of dealing with thousands of different types of requests from residents and prospects. - Your team can adjust its behavior based on changing situations. - Your team can interact with your other software systems. Large language models can't do any of these things on their own. ## So How Can We Use AI? In a nutshell, your teams have more than just the ability to use language; they have layered resident, property, company, and industry knowledge. They also have job experience, awareness of industry best practices, process knowledge, and decision-making experience. To bring AI into your team, you need to turn to industry-specific AI solutions. Solutions like Adam, where a team of AI & real estate specialists work solely on layering real estate knowledge, processes and experience on top of more general models, to create a tool that compliments the work of property managers, allowing them to prioritize the tasks that are most important for their residents and team. The models that OpenAI, Facebook, Google, and others are pioneering form some of the building blocks, but to effectively supplement your team with an AI teammate, you’ll need to partner with domain-specialized R&D companies. There’s no doubt that in the next ten years AI will impact your business. By partnering with specialists, you will be able to look past the hype and unlock the power of AI. ## Travtus at OPTECH URL: https://www.travtus.com/resources/travtus-at-optech Date: 2022-11-23 Summary: Once again, the NHMC’s annual gathering at OPTECH provided insight into the real estate industry. With a long list of housing providers, suppliers and startups attending, the schedule was - as always - comprehensive, covering topics as wide-ranging as AI, centralization, automation and resident experience. Once again, the NHMC’s annual gathering at OPTECH provided insight into the real estate industry. With a long list of housing providers, suppliers and startups attending, the schedule was - as always - comprehensive, covering topics as wide-ranging as AI, centralization, automation and resident experience. ## AI was in the Spotlight The potential of AI and machine learning to optimize operations and drive executive decisions was a trending topic this year. Industry doubts around gathering enough clean data remain, but when possible, owners and operators of all sizes are noting that AI can add value throughout their business, from operational benefits to intelligence-led, top-line decisions. ## Centralization is Happening Centralization is moving from being often misused industry buzzword to operational strategy, with marketing and leasing leading the way. The centralization of maintenance and back office staff into multi-community clusters is next on the agenda, driven by the problematic churn of Multifamily employees (more on that later). UDR’s Josh Gampp summed up the all-or-nothing organizational challenges of centralization “You have to be bold. If you want to move toward centralization, start by going to your CEO. If you don’t get their buy in, give up.” ## Streamlining Tech is Easier Said Than Done The evergreen challenge of managing complicated tech stacks is still causing consternation across the industry. Stretched by decreasing tech budgets, multifamily operators struggling with the demands of running and integrating various software platforms are becoming increasingly focused on streamlining their technology stack while still requiring specialized solutions for their individual requirements. This streamlining means favoring platforms that can provide a menu of solutions or agile companies that can respond to off-menu needs efficiently. ## Properties Need Managed Services “Managed services” are coming to the fore, with a single network providing residents connectivity and supporting advances in intelligent buildings. But as technology marches ever forward, developers and operators are struggling to future-proof communities. Experts favor a future-friendly approach to upgrading existing properties using conduits and pathways to allow easy upgrading in the future. ## Staff Problems With the average tenure of an apartment industry worker reported to be at a shocking 2.5 years only, the industry is still struggling with retaining talent. It seems good old-fashioned word of mouth is a key recruiting method, with employers who put effort into employee well-being finding it easier to recruit and retain talent. ## We’re Backed by Our Customers: Travtus Raises Seed Funding Led by RET Ventures URL: https://www.travtus.com/resources/were-backed-by-our-customers-travtus-raises-seed-funding-led-by-ret-ventures Date: 2022-11-02 Summary: We’re pleased to report that Travtus has raised seed funding led by RET Ventures, an industry-backed venture fund focused on single and multifamily real estate technology, allowing us to take Adam, our everything teammate, to the next level. We’re pleased to report that Travtus has raised seed funding led by RET Ventures, an industry-backed venture fund focused on single and multifamily real estate technology, allowing us to take Adam, our everything teammate, to the next level. Adam shows residential owners and operators the benefit of one single powerful machine learning platform to handle multiple job functions, including that of a data analyst, agent, resident experience representative, and teammate for everything community management. > “As we continue to penetrate the market, we are thrilled to have a trusted partner like John Helm and the RET Ventures team, who are so deeply entrenched in the industry. This investment will give us the capabilities to ramp up our sales and marketing efforts and expand our product development – allowing us to deliver an even more robust product to improve property management for multifamily and beyond.” Tripty Arya, CEO of Travtus. With this new funding, we'll continue making it easier for community managers to scale operations and deliver the best resident experience. Read more about this exciting new stage in our journey: [Travtus Raises Seed Funding Led by RET Ventures to Optimize Machine Learning for Property Management | Business Wire\]](https://www.businesswire.com/news/home/20221101005900/en/Travtus-Raises-4-Million-in-Seed-Funding-Led-by-RET-Ventures-to-Optimize-Machine-Learning-for-Property-Management) ## Travtus + Cortland URL: https://www.travtus.com/resources/adam-cortland Date: 2022-10-31 Summary: Cortland uses Travtus products and labs to build a resident-first operating platform which is designed to improve visibility into onsite operations and resident needs. #### Adam Makes Every Customer Interaction Count [Cortland](https://cortland.com/) is a real estate investment and property management brand that focuses on delivering resident-centric, hospitality-driven apartment living experiences in the US and UK. It currently owns and operated over 60,000 apartments and is a prominent member of the Top 50, NMHC list of owners and operators. Cortland’s growth is fueled by a resident experience platform rooted in innovation, service, communication, and consumer insights. Cortland is an early mover in recognizing the benefits of Adam’s natural language models to inform their operations and service to their residents. #### Adam at Cortland Adam is an artificial intelligence platform for property operations. It is not a chatbot or a software system. It is a collection of domain-specific machine learning models combined with layers of knowledge about real estate operations, the property, and users to create actionable tasks. It’s an operating system to streamline property operations with automation and front-line intelligence.  Adam was recruited as a member of the management team for select apartment communities within Cortland’s Atlanta portfolio with the goal of delivering front-line intelligence to enrich the level and consistency of direct resident service and communication. Adam provides the Cortland team with data insights using high-quality, natural language models that recognize patterns in resident and asset behavior. #### Using Technology to be More Human Cortland’s focus on their residents’ needs presents the unique challenge of consistently elevating the human interactions of customer experience while also creating efficiency and consistency within that experience. Residents can use an online portal to make payments and raise service tickets. However, that platform can often become a barrier to the essential feedback and authentic interactions that accompany the voice of the customer. It’s these insights that help maintain a core customer focus and evolve service and product offerings.  At Cortland, Adam helps the onsite team maintain the personal aspect of conversation with residents through text while empowering the team with data to help inform improvements to resident retention and the lifetime value of a customer. Adam’s unique proposition of customer self-service through tech while providing key behavioral data allows for the efficiency and cost benefits of text while minimizing disruption to the ways customers interact daily with the Cortland teams.   #### Using the Advantage of Scale Cortland’s goal of creating a forward-thinking resident experience at scale is not an easy task. Property management is complex due to the breadth of scope, which includes hundreds of questions and tasks that the onsite teams have to handle on a daily basis. The right technology solution can assist in making these tasks more manageable while using the learnings from across many portfolios to create successful brand expectations. By recruiting Adam, Cortland is making each resident interaction with Adam count. The platform has not only helped manage the teams’ resident inbound inquiries but has also added a key level of visibility into their onsite operations and resident behavior. Cortland’s management team can now identify opportunities to improve processes and engage in proactive communication to unlock better customer service – which ultimately helps them become the apartment living brand that people love and trust. ## Are Multifamily Properties Understaffed? URL: https://www.travtus.com/resources/are-multifamily-properties-understaffed Date: 2022-10-30 Summary: Why is the onsite staff always recruiting and what do residents need them to do. Adam explores the real reason for increasing headcount. The recent report by NAA published that for every group of 45 apartments at least one full-time employee has to be added to the payroll. Labor shortage and talent retention remain the top concerns for owners and operators. At the same time, the need for business intelligence for improved remote working is becoming mission-critical. According to the report, the need for on-site staff has increased due to bad resident reviews and increasing expectations. But the data from the ground gives us a more detailed and nuanced picture of the problem. Domain-specific natural language models enable us to examine resident and prospect conversations at scale while drawing both automation and insight which would normally be invisible to multifamily owners. In 2020, ADAM, our proprietary AI platform for multifamily operations, had over 21,000 unique conversations with both prospects & residents across the US. Here is our aggregated insight from on-site operations. ### **Timing matters** Residents demand support as and when they need it. The volume of interactions with residents surges by 2.5 times on the first two days and last two days of the month. This trend is in fact predictable and repeats consistently every month. Also, over the course of the year, 36% of the interactions and 46% of leads came from outside of the typical office hours of 10AM to 6PM. This variance creates a major staffing challenge. In order to do their jobs, the staff interacting with the residents need to be not just knowledgeable but also scalable by a factor of 2.5 . This pattern also results in task requests coming in like a deluge and creating backlogs for the rest of the month. Variance is less acute for the leasing staff, who see a steady flow of prospects throughout the month. However, the months of June, July, and August had 5 times the number of prospect conversations that December and created 3.5 times more leads than any other period of the year. ### **A very long tail** The volume and importance of maintenance requests are being overestimated and all other topics are in fact underestimated. The largest cluster of interactions making up 42% of all resident interactions involved people asking questions about the property or policies of use.  This cluster itself breaks down into over 700 different types of questions as specific as ‘how to use a gate’ to ‘how to use the app’. Only 14% of the resident interactions were related to property maintenance. Payment-related inquiries contributed to 10% of the volume eventually leading to a very long tail of various unique topics ranging from deliveries, complaints, waste disposal, promotions, move-ins, move-outs & beyond. The challenge that long tails create for on-sites teams is that they constantly feel that they are wearing multiple hats and are being pulled in too many directions. At the same time, residents feel like they are not getting a consistent experience or a response to their queries. ### **Process gaps** The use of natural language models allows visibility to the on-site data with precision but it also exposes gaps in processes with which the site staff often improvise. In 25% of all conversations, the interaction had to be passed to someone on-site in order to figure out what needed to be done, despite the models predicting the expectation of the resident correctly. These topics included items, often repeated, which do not have a defined process or system to be inputted into. This is not a problem of interoperability of software but of missing processes that come from the historic opacity of the sites. The onsite teams will often mention burnout and stress as the main factors for talent retention issues and turnover. In the absence of organizational learning, much knowledge is lost with staff departure. This makes it challenging to set performance metrics for teams where the SLA’s and KPIs are often arbitrary or based on a very small part of their role. ### **Finding micro-trends** The long tail challenge of property management is addressed well by the opportunity of finding early micro trends by leveraging machine learning techniques. The on-site teams perform better with intelligence support. However, this requires the adoption of very high quality and specialized models. Some examples of insight gathered with micro-clusters include a relative increase in prospects wanting in-unit washers & dryers. The sharer lifestyle was also highlighted in 2020 with more prospects asking about shared kitchenettes in the community and quiet places to work or take exams. Microtrends were also observed with more residents asking to use the management office printers & faxes. Conversations about rental promotions were had in equal measures by both residents & prospects, indicating a new pipeline for apartment transfers. In conclusion, multifamily properties are not understaffed but they are dealing with a resource allocation problem which includes a variance in both volume & variety of tasks. It is the reason they always seem to be overstretched and under-resourced. Using cutting edge machine learning assistance can not just reduce the challenge of recruiting and training. It can also create a base for a consistent and improved customer experience. ## Are resident portals effective? URL: https://www.travtus.com/resources/are-resident-portals-effective Date: 2022-10-30 Summary: Resident portals are an excellent placeholder to direct basic workflows but in order to truly get value from technology, you need solutions that can account for and react to the complexity of managing properties. The internet is full of articles on the benefits of onboarding a resident portal. All publicly available content about the effectiveness of resident portals is either published by the SEO teams of providers of resident apps like Zego, Yardi, RealPage or articles in the industry press again including the quotes from the vendors themselves. If you search “are resident portals effective’, there is no answer to the question. Just pages and pages about the benefits of a resident portal. The adoption rate on resident portals is at an all-time high. Yet, the industry continues to have one of the lowest customer satisfaction scores and employee turnover for onsite staff is consistently high. The correlation between the adoption of technology and better resident and employee experience is not yet proven. ### **Capturing Nuance in a Portal** The main issue with the limitations of the resident portal is that it is envisioned as a hub for self-service. The goal is for the residents to log in and complete a transaction without a phone call or a visit to the onsite office. In the most rudimentary portal, this includes the ability to pay rent and to create a service request. The more sophisticated and ‘feature-rich portals include additional automation like bookings, package notifications, and community news. However, when Adam engages in a conversation with a resident over time, we learn that these self-service items make up a very small proportion of resident goals. The greatest challenge for operators is the very [long tail of the nature of resident interactions](https://www.travtus.com/insight/are-multifamily-properties-understaffed). This means that residents need help with a large variety of things but in relatively small numbers. This makes it very difficult to break down the self-service options in a way that can have a material impact on the support volumes. Ironically, after leasing and maintenance-related interactions, the most popular topic of conversation is [technical support](https://www.travtus.com/insight/technical-support-is-the-fastest-growing-overhead-for-rental-operators). ### **Losing frontline intelligence** The adoption of resident portals has also added a barrier to direct customer feedback. While most operators look to reduce the interactions between their residents and onsite staff, they also risk losing the front-line intelligence that resident interactions generate. The restricted screens of a portal don’t allow for exception handling and lose the nuance that human interactions pick up in day-to-day operations. Similarly, the codified FAQ section often results in boilerplate assistance which only adds to their frustration. The business case for onboarding resident portals lies in the hope that self-service options reduce the demand for man-hours on the onsite staff. But resident portals work mostly as landing pages to a handful of tasks and are often bypassed for another medium of least resistance without logins. In the situation where a resident is offered multiple channels of communication, 7 out of 10 residents choose to bypass the resident portals because they believe that it can’t support the exception to the situations designed into the portal. ### **Solving the Problem** This does not mean that technology does not have a place in multifamily operations. In fact, it builds a case that the residential operator needs not just technology but cutting-edge solutions that may not have the same applications in other industries. Behind all the screens and user-friendly UX of the portals and apps, lies a database that is populated through workflows. The operators need to focus on triggering these operations with a more suited technology of Natural Language interaction where residents are able to communicate nuance and at the same time employees are able to focus on the tickets that represent true exceptions. Resident portals are an excellent placeholder to direct basic workflows but in order to truly get value from technology, you need solutions that can account for and react to the complexity of managing properties. ## CBRE Talks to Adam About AI Innovation URL: https://www.travtus.com/resources/cbre-talks-to-adam-about-ai-innovation Date: 2022-10-30 Summary: CBRE asked Tripty Arya of Travtus, about the potential impact of Artificial Intelligence, alongside the complexity of implementing it in the commercial real estate world AI has the ability to automate processes in a facilities management context, understanding equipment usage metrics on an ongoing basis and providing automated, actionable insights as a result. This, ultimately, can drive down energy and maintenance costs while also ensuring the smooth running of services for building users throughout the day. We asked Tripty Arya, Founder, and CEO of Travtus, about the potential impact of Artificial Intelligence, alongside the complexity of implementing it in the commercial real estate world. ## Customer Experience in Real Estate needs Tech URL: https://www.travtus.com/resources/customer-experience-in-real-estate-needs-tech Date: 2022-10-30 Summary: Powered by open data, big data tools consolidating information from across your business and AI tools that offer customer centric experience Financial services and retail industries are obsessed with a single customer view. Having a holistic view of your customers and being able to provide them great customer experience – with good reason – you hate having to answer the same questions 1000 times! The last 15 years have been a relentless drive for better customer experience, increased sales, and better customer retention through data. So why is the real estate industry so behind, why do we not share this obsession with consolidating our data and leveraging its value? How can we learn from other’s mistakes and find a new source of competitive advantage? ## **The birth of single customer view** In 2004, the dot-com boom (and crash) had brought online experiences into general view, Amazon & eBay were stealing market share from traditional retail. Online payment standards brought to trust in online experiences. The retail world took notice and the ability to provide consistent customer experiences; online, in person, or on the phone became the key to winning new business and retaining customers. Banking soon followed suit in this obsession, however, complacent in their dominance & regulatory protection and constrained the baggage of 30 years of legacy software ( one system for credit cards, another for current accounts, savings, acquisition…) progress was slow. The same CTO’s that had implemented these systems turned to traditional “proven” technologies to integrate into a search for efficiency and that all-important single customer view. An entire industry of consultants, offshore development, and expensive tools emerged. > “Globally $billions have been spent on incremental improvement, customer experiences that still lag other industries and all the while Fintech has innovated quickly posing a real and significant threat to the established.” It’s not that banking failed in its goal, but it’s been a long game of constant catch-up and at an incredible cost. ## **So, what about Real Estate?** Real estate today finds its self in a similar position as the banks in mid-2000s. Examine a typical property management company, they have tools for accounting, leasing, marketing, work order management, inventory management, renovations, inspections, asset management, and the list goes on. They have slowly migrated from paper to database tables to perform specific functions. With the emergence of proptech the “forward-thinking” companies have adopted more tools, enabling them to digitize more processes while at the same time creating even more silos of information, MOUNTAINS OF VIRTUAL PAPER, and more dashboards than anyone will look at. The real estate software industry has only perpetuated the problem. Incumbent providers of accounting and building management solutions, less nimble than startups, charge as much as $25,000 per year to their existing customers to access their own data and integrate with other tools. These large companies form partnership schemes that discourage competition and pose barriers to industry innovation while obstructing their own clients from becoming more integrated and efficient. As the real estate industry wakes up to the potential of data it finds itself in a situation as bad, if not worse, than the banks. If we are not careful millions will be spent on patching together disjointed experiences in the drive to offer customers the experiences they have come to expect. Costly projects to deploy yesterday’s technology will take 2 years and by the time they go live we will again be playing catch up. > “We must not fall in to the same traps as the banks, we have to learn from their mistakes.” We need to look at the new tools and solutions available to us. Tools built around big data, machine learning, and AI. But for this, the industry needs to first accept and embrace the idea of data collection & sharing. As an industry, if we keep looking to the traditionalists for solutions and more “joins” on our tables, we won’t have any real progress. ## **Imagine a new world…** Tenants’ preferences are remembered and mapped to your properties, leasing policies, local amenities. You can offer suggestions from within your portfolio for a switch in the lease to a new apartment closer to their office. Rent pricing can now be linked to their history of requests and complaints. A single customer view that can help you focus your efforts on retaining good tenants, offering them bigger or newly renovated properties to suit their lifestyles. Enabling you to identify high-risk tenants that you need to manage more closely to recover rent. It doesn’t stop there. This structure applies to your assets, to your vendors, employees, company policies, and the overall health of the portfolio. These capabilities are within our reach now! Powered by open data, big data tools consolidating information from across your business, and AI tools that offer customer-centric experience. Look to the future to drive your business not the past! ## Diversification Is Not Innovation URL: https://www.travtus.com/resources/diversification-is-not-innovation Date: 2022-10-30 Summary: Real estate needs true investment in to people who are risk takers and outliers. It is from here that radical, head spinning change will come. In 1955, 500 companies featured on the first “Fortune 500” list. Only 12% of those companies survive today. Sixty four years is not an eternity. We carry memories and lessons from the wars of another generation but often forget the changes that defines survival within our own. Moreover, the velocity of change has accelerated with time. A report released by “Innosight” in 2016 confirmed that > “The 33-year average tenure of companies on the S&P 500 in 1965 narrowed to 20 years in 1990 and is forecast to shrink to 14 years by 2026.” This means that half of the companies which are on the S&P 500 today will be replaced within a decade by a new entrant. The ability to change and adapt to new trends is not a matter of growth but that of survival and the real estate industry is slowly coming to this realization. A new movement called “ Proptech” has become the buzzword for the industry. By definition, Proptech is the intersection of the property and technology built to propel the industry forward. But the label has now expanded to include any new changes that are seen in the traditional model of real estate investing. From Co-working spaces to photographing drones, the entire gamut of any changes to the industry is now collected within the larger context of “Proptech”. The acceptance of Proptech into the industry is a welcoming phenomenon but it has created a new dilemma. In order to achieve this acceptance and short-term wins, the direction of innovation is being held back due to the bias that the industry has towards innovation. ## **Diversification is not Innovation** In the 1980s management was being turned on its head. New degrees and new skill sets were being developed. It saw the birth of new ideas like Six Sigma & Matrix Management. And along with it came the concept of Portfolio Theory. The era was one of managerial bravado. Entering into the nineties, KKRs buyout of Nabisco represented the epitome of the change in the attitude towards the impact of management. This is also the same period when the Harvard Business Review came out with an article by Professor Ralph Briggadike called “ The Risky Business of Diversification”. This report had an impact on the new leaders of yesterday. It used a data-driven process to attest to the ideal time that must be given to a business in order for it to succeed. > “New Ventures need on the average, eight years before they reach profitability”. Under the guise of diversification, the concept of corporate venturing was born. Corporates began to seed new product lines and the idea of “new” was immediately linked to diversification. However, the success of a “venture” was founded on the idea that products or services were being made for the existing market while utilizing the same distribution channel. The sunk cost discounted; the marginal impact of diversification was worth the investment in the “new”. This was also the period when Real Estate as an institutional asset class was maturing. The industry took the view of diversification in order to address two types of risks: Market Risk and Technical Risk. Portfolios over time were diversified by asset class. Innovative financing structures were set up like the creation of securitization. The market eventually drove to changes in lease structures and terms. However, the industry has always shied away from investing in combating technical risk. The birth of Proptech was the first step towards addressing technical risk but from the view of incremental efficiency gains. The examples of WeWork & Compass are narrated as the success stories of Proptech. But neither is an example of the diversification strategy of incumbents. They are a product of the greatest advantage of a startup which is to start from scratch with the relevance of today without the baggage of legacy. These are not startups that hustled. They conquered and will continue to do so. However, the attitude of incumbents today towards Proptech continues to apply a proven concept to their existing distribution lines. They don’t recognize that the greatest gain from innovation comes from creating distribution lines or markets that have not yet truly existed. > The application of technology is being restricted to arguments of efficiency and value add. Innovation is skipped in order to get instant gratification and the end result is a market full of noise and new dashboards. ## **Venture Capital is not Private Equity** In 1980 McKinsey was commissioned by AT&T to forecast the cell phone penetration in the US for 2000. Based on the existing metrics and intelligent projects, their prediction claimed a potential market size of 900,000 users. The true number in 2000 was 109 Million users. It is this uncertainty that is the very basis of Venture Capital. In 2013 when Blockbuster closed its last retail store, it was an indication that Netflix had won. When Netflix started delivering DVDs in the mail instead of a quick run to the local Blockbuster, the difference between the “David” and the “Goliath” didn’t seem very large. All they did was set up a website and some postal envelopes. At least this is probably what Blockbuster’s board assumed. Actually, I am sure that the board had not heard of Netflix till it was already too late. It was too late when Netflix without its baggage started innovating on content delivery and the content itself. It’s not like Blockbuster did not follow suit or set up websites and streaming. But by then, they were already playing catchup. The state of incumbents in Real Estate is the same. They are looking at the innovations around them and grasping at first just catching up. In order to do so, there are many real estate companies investing in “Proptech” funds or setting up their own Venture Capital arms. But how does an industry that has always structured itself with Private Equity investment change everything they know to become venture capitalists? The real answer is that they don’t. Instead of looking for Netflix, they are constantly looking for solutions they can “invest” in replicating the Blockbuster model of catchup. There is much capital that is being raised for Proptech with the single aim of “futureproofing” incumbents. The mandate for most of this cash is to provide technology to preferred LPs who can apply and direct the products of companies to work as outsourced solution providers. The real estate industry understands Private Equity and is applying it to their new-found status in Venture. In Private equity, you start with numbers. These numbers represent success, cash flow, and a clear tactic. The job of private equity is to optimize. However, Venture Capital does not start with numbers. It is based on people. Its successes are often linked to failures. > The challenge with private equity entering venture capital is that the innovators are being forced to behave like participants in a six sigma process instead of letting loose and fly or fall. ## **Embracing the Unexpected** In May 2015, Jack Ma addressed students in South Korea and outlined his “advise” for the progress of an individual. He broke life down into decades and assigned purpose to each decade. Before your twenties, life is about learning. Before your thirties is about following a good leader who you can respect and learn from. From Thirty you must find purpose and set the groundwork for your life’s work. Your forties are to follow all your training and passion to do what you know well. The fifties is when you work with the younger generation and invest in them. And the sixties are about spending time on yourself. While we can all relate to this from a personal perspective, we should be applying the same view to the lifecycle of companies. We could compare the incumbents of Real Estate as going through a mid-life crisis. They are the equivalent of the 50-year-old successful man who is holding on to his methods of success. They are trying to hold on to their relevance, trying to teach younger companies to do things their way but by using new tools. Instead what is needed is encouragement and investment which allows for the next generation to move forward. The strategy of supporting proptech with the bias of diversification & private equity is delaying the inevitable. > Real estate needs true investment in to people. Creative people, risk takers, outliers. People that are building a new future for the industry with no strings attached. It is from here that radical, head spinning change will come. ## Ensuring Success of a Property Acquisition within the first 90 days URL: https://www.travtus.com/resources/ensuring-success-of-a-property-acquisition-within-the-first-90-days Date: 2022-10-30 Summary: By focusing on field intelligence, talent retention and effective change management, each acquisition can bring value that can position an organization for further growth The first 90 days of any acquisition are crucial to the success of the investment. Turning the purchase of a property into a success requires not just tremendous skill but also a culture of agility and learning on the ground. The pandemic has created a growing appetite for Value Add strategies as more operators are needed to help capture the inflationary benefits of the growing rental market. This poses the traditional industry problem – how do I scale. **Defining the challenge** Most operators struggle to deal with the change management of taking over a new property despite having done it multiple times. The challenges usually are linked to communication to residents and training of on-site staff. More often than not, the asset is purchased with the onsite team in place. You have both new customers and team members to adopt new processes and absorb into your “way of doing things” while physical changes are mostly minor with signage changes and evaluation of future renovation pipelines. The acquisition period is one of additional support with phones ringing off the hook while the staff is being trained by the new management company. The data collected with Adam shows that the resident support requests double in the first month of acquisition. The operating team needs to hit the ground running while maintaining their brand value by providing a good first experience to their newly acquired renters. ### **Prioritizing Frontline Intelligence** The first few months can produce the platform for a very successful value add strategy by doing one very important task – listening. If you are entering a new market with the acquisition or have been purchasing the whole street, every community is unique. It has its own challenges which may be linked to the physical condition of a property, the service offering, or resident needs. The exercise of engaging in gathering field intelligence is often missed in the chaos of management take over. It is critical to learn about the new asset as quickly as possible by using all tools necessary to pick up trends. Operators who focus on a data-first approach to strategy are agile in prioritizing the pain points first and build trust from day one. > “We recruited ADAM within the first month of acquiring our largest asset in Denver with the only purpose of learning more about our new property and residents. We were able to take stock of the situation within days instead of waiting for months of settling in.” ### **Making changes fast** Most operators try to continue operating with the status quo and trickle-down change over months. Any changes including the addition of technology or reorganizing of teams should be done early on in the acquisition. The changes in personnel or service offerings should be introduced within the first 90 days in order to get buy into integration objections of the operating platform. When every property is managed in its own inherited format, it only creates a burden on the operating platform it is now a part of. Properties which integrate into the new owner’s platform within the first 90 days are able to benefit and contribute to the overall fund or platform return quickly. > “When we purchased this portfolio, we wanted to minimize change and yet bring improvements. Adam was an easy decision because it helped us bring a lot of change without it impacting the status quo for the residents or the teams.” ### **Managing Recruiting and Talent Retention** The first few weeks of an acquisition can stress the inherited staff. They are meeting a new employer and at the same time dealing with additional support with phones ringing off the hook with residents asking about the changes they should expect. It is also a time when they are booked into training sessions with their new employer and are unable to focus on their day-to-day operations.  Over 40% of new operators see staff turnover within the first 90 days of an acquisition. As of July 2021, there are 70,000 active online job openings in property management in the United States. It is important that the new operator comes to their new acquisition with a set of tools that can impact the team’s productivity without additional chaos and pressures of training. > “The day of the first rent payment after we took over the property was hell. We had so many calls asking about how to make the payment and what payment types we would accept. Having Adam handle these in that week was like having extra help in the office. ” **Productivity, people & profit is a value add strategy** An acquisition is a time of great celebration although it usually has a very little instant pay-off. By focusing on field intelligence, talent retention, and effective change management, each acquisition can bring value that can position an organization for further growth. However, an acquisition that is operated in a reactive way will be a drain on the resources of the larger organization and hold back its growth. ## Technical support is the fastest growing overhead for rental operators URL: https://www.travtus.com/resources/technical-support-is-the-fastest-growing-overhead-for-rental-operators Date: 2022-10-30 Summary: Did you know that 8% of all interactions with residents and prospects are for technical support? As more rental operators adopt tech, assistance and reporting issues with tech have been steadily increasing. ### **Most rental operators believe that maintenance & leasing are the two main functions of property operations. However, maintenance only makes up 15% of all interactions, and leasing can range between 20% to 11% depending on the month. But did you know that 8% of all interactions with residents and prospects are for technical support?** As more rental operators adopt tech, including leasing applications, resident portals, package lockers, and smart gates, the outreach for assistance with using the tech or reporting the issues with tech has been steadily increasing. It is important to note that while most technology is supported directly by the vendor, the end-user first reaches out to their property manager or landlord to support any technology provided to them.  Most of this support, today, is handled by an ad-hoc manager and often lost in a three-way conversation between operator, customer & vendor. The four biggest drivers of technical support are: ### **Data Inconsistency** The real estate industry is notorious for bad data quality and its lack of open infrastructure for data to flow freely. A large number of support tasks are related to occupants and tenants not being registered across platforms and the use of phone numbers and emails which don’t match users and systems. This type of support usually falls back to the property or leasing staff since it is not something that can be resolved by the vendor’s support task. ### **Disjointed User Experience** Residents and prospects recognize the brand of their landlords but not that of ‘proptech’ solutions. ‘Bolting’ multiple software solutions together without designing the user’s journey impacts their experience of your brand. The addition of technology as a layer between direct interactions has also shifted the user’s need for support from – _“I want to”_ to _“How do I”_. The user is looking for support on what tools they have to interact with to achieve a specific goal. These interactions fall to the property teams to assist the users with the knowledge of the platforms to trigger a required process. ### **User Error** The most popular example of user error is seen with package lockers when delivery service professionals scan the wrong package or misplace an item in the wrong locker.  In the case that a resident raises a ticket directly with the package locker vendor, these support tickets are often sent back to the on-site teams to resolve on the ground. Most user error support tasks for cloud-based software can be escalated to tech vendors but often result in the residents wanting to use an alternative method. A great example is users roadblocked trying to make a payment before the due date, wanting an alternative method to make the payment instead of waiting to resolve the error. ### **Technical Bug & setup** The smallest contributor to technical support concerns is an actual bug or a problem with the configuration of the software or hardware.  These bugs are often reported by the residents or prospects but have to be formally raised as a ticket by the operator with their technology vendor. The adoption of technology is essential for rental operators to stay nimble and relevant to the future, however, technical support is crucial to the overall success of the strategic benefits of technology. When your customers run into a problem, they call you and not your vendor. These questions need to answer with technical expertise. A very common reason for attrition of on-site staff is the lack of training or tools to get the job done and adding technical support to the burden of the ‘long tail’ makes the job more unstructured. ### **Technical support is well suited to a centralized function for any operator who wants to increase productivity and scalability. It is also important to consider the cost and volume of support when selecting rigid industry platforms as opposed to moving to non-industry-specific solutions that have additional flexibility and community for support.** ## Towards a better employee experience for onsite teams URL: https://www.travtus.com/resources/towards-a-better-employee-experience-for-onsite-teams Date: 2022-10-30 Summary: The most crucial aspect to reduce employee turnover is not just training but providing the teams with the tools and support to truly succeed at their jobs The high turnover rate of property management staff in multifamily housing is an established fact. At a 33% average rate, it is a major contributor to the inefficiencies that the industry faces. The high cost of replacement of an onsite employee has now put a sharp focus on the employee experience programs at every large multifamily operator. ### **Understanding the Job** A job by definition is a task or multiple tasks that come together to achieve a goal. Many job descriptions for onsite teams may refer to the goal of assisting residents or prospects, they are less forthcoming in the descriptions of the task. And the reason for this is within the inherent nature of the job that property management is an industry that has a [very long tail of tasks](https://www.travtus.com/insight/are-multifamily-properties-understaffed) and yet needs excellent knowledge for the completion of this task. As a result, the lowest denominator role in the company which is often the onsite role is in fact the most entrepreneurial in nature. It requires someone to be able to problem-solve and be empowered to take decisions. This is a very rare quality to find. As a result, while there are 70,000 online job listings for property management roles in America today, the number of entrepreneurs who are a good fit for the job is scarce. At the same time, operators want to capture more front-line intelligence to also support their platform. This means that often the entrepreneurial onsite role is burdened with process and data capture responsibilities and as a result creates a conflict in the core competency expectation for the team. ### **Separating data from operations** \*\*\*\*It is important that the onsite teams are focused on the job that provides the most satisfaction and also provides your clients the best service. And this means that the responsibility of data entry and process are separated from the onsite teams. Using technology platforms that leverage Natural Language Processing like Adam, allows your residents to engage in the most natural way possible while creating tickets for problem-solving for the onsite teams. Any items that are part of a process flow can then be automated with the use of technology without distracting the teams. It is also important to provide the onsite problem solvers with the time to think. A task list can help the team discuss a problem and also have all the information at hand before speaking with a resident with a solution. ### **Provide Intelligence to the field** Most properties aggregate information on the site level and report to the regional teams. Operators who want to retain the best talent in the industry will instead provide the aggregated field intelligence to support their teams in the field. Data capture through portals, digital employees, or communication should be analyzed and aggregated to highlight trends that the onsite team can miss due to their focus on completing a task and the day-to-day. The most crucial aspect to reduce employee turnover is not just training but providing the teams with the intelligence support to truly succeed at their jobs. ## Transforming Customer Support to Customer Success for Multifamily Operators URL: https://www.travtus.com/resources/transforming-customer-support-to-customer-success-for-multifamily-operators Date: 2022-10-30 Summary: Resident support should not be treated as the unwanted consequence of having a client. It is a window into an opportunity to be better and gain a competitive advantage. As a multifamily operator, you have done your best at optimizing the occupancy of your building. You may even be improving the tenancy stock at the property and investing dollars to attract a new demographic. You have found your product-market fit and moved in the right pipeline of the tenant. You are now looking towards your next acquisition or even the next move-in cycle. But at the property, the phone is ringing. The emails are getting clogged and you feel like the team in always recruiting. **The long tail problem** ADAM’s data from over 30,000 conversations help us identify the challenge for operations. Property management has a long tail problem. There are many low volumes, hard-to-find topics of support which are come in at varying demand cycles, creating backlogs for support. But, your customers live in a world where Amazon and Apple set the standards for customer attention. As a result, the relationship between the customer and the operator is often fraught with friction. In any industry, as companies scale, they implement processes and try to adopt automation. Property Management is no different. Most reputable operators have at least a basic resident portal. The valued customer signs a lease and becomes a resident. They pick up their keys and some information about using technology to set up payments and the resident is now put on autopilot. At least that is the hope. The resident’s experience is now restricted within the confines of the buttons of a resident app or the unanswered calls of the onsite office. **Dissatisfaction is feedback** Most property operators are missing the big picture. The key to understanding your product, its performance, and also its success is to not avoid customer dissatisfaction and think of it as an item of reactive support, but to engage in the ‘long-tail interactions to bring customer success. Relying on call centers and portals removes all feedback loops which is the true intellectual property and learning of your business. The key to building a scalable and profitable product is to understand what your customer wants. Customer support should not be treated as the unwanted consequence of having a client. It is a window into an opportunity to be better and gain a competitive advantage. In all industries that have engaged in customer success strategies, the rewards have been great. They have seen increased profitability due to both additional revenue and cost reductions. It also creates happier and more engaged employees while organically building a brand. **Frontline Intelligence without the effort** A professional property operator will invest in creating a solid infrastructure that allows customers to get more from interacting with you. The challenge is always the quality of data, interaction, and the cost associated. The adoption of a conversational machine learning platform is the first step towards bringing front-line intelligence into your board rooms. The insight can then help create opportunities for both additional offerings and savings and delivered in a natural conversation. The adoption of the right technology can help create a real relationship with your customer without throwing more bodies at the problem. ## Want better leads? Turn the funnel upside down URL: https://www.travtus.com/resources/want-better-leads-turn-the-funnel-upside-down Date: 2022-10-30 Summary: The operator need not cast a wide net but instead turn their funnel upside down and focus on their existing residents to build very targeted messaging and campaigns to attract a truly qualified lead. Marketing a multifamily property used to mean putting up a sign on the façade and putting out a listing on the syndicated feeds. However, a lot has changed as these properties have evolved into an institutional asset class that is professionally managed and optimized for returns. Professional marketers are bringing to the industry best practices from other industries. Apartment descriptions online are being treated like product descriptions. The images and tours are now digitized professionally to reduce the length of the sales cycle and improve the quality of the lead. Some sophisticated groups are even looking at retargeting campaigns and online ads to drive traffic to their own sites in the first steps towards brand development. As competition increases, the product differentiation is no longer restricted to just location and amenities but also includes brand experience. ### **Your resident can tell you what the prospect wants** The typical marketing team at a multifamily operator aims to build awareness and drive leads through a traditional sales funnel which culminates with a lease commitment. The operator communicated the availability of an apartment over a variety of platforms, often listing sites or through broker networks. These generate views and collect visitors which fill the top of the funnel. As your preferred resident gets closer to paying you rent, more and more of these leads drop out. The resident is the small group left at the very bottom. However, the multifamily industry has an advantage over most other industries. It is fragmented and local. The largest of operators at any point is looking for no more than 40,000 people over multiple geographies to occupy their product. As a result, each operator need not cast a wide net but can in fact turn their funnel upside down and focus on their existing residents to build very targeted messaging and campaigns to attract a truly qualified lead. The ‘upside down’ approach allows operators to focus on their residents and learn from their experience. This approach has a few benefits. Firstly, the resident focus keeps vacancy rates low and increases the long-term value of your customer. This means that every dollar spent on acquiring the resident has a higher rate of return. The second benefit of this approach allows marketing teams to truly simplify the messaging and product offering. Every conversation with a resident is an indication of what they like or dislike. If the goal is the attract a similar profile resident, the field intelligence gathered through support and usage data also informs the value proposition that can drive marketing campaigns that appeal to the same demographic. ### **Field intelligence is a tool for marketing** The upside-down funnel comes with its advantages, but it also requires a culture that puts support and field intelligence at the very core of the business. The business needs to acknowledge their inputs and showcase a genuine commitment to providing an easy and enjoyable living experience. This includes easy access to information as well as prioritizing operational efficiencies which can cultivate a great resident experience. It is critical to leverage technology to reduce the burden of balancing good customer experience with churn and employee strain. The field intelligence gathered by the use of natural language technology by Adam assists not just in reducing the workload for onsite teams but also highlights the trends and needs of the residents of each individual community. The appeal of every community is unique and the messaging for its marketing needs to be targeted to that unique appeal. ## What does Latch’s public debut mean for Multifamily Owners? URL: https://www.travtus.com/resources/what-does-latchs-public-debut-mean-for-multifamily-owners Date: 2022-10-30 Summary: When Latch announced its merger it was offered an Enterprise Value of $1.05 Billion on reported Net Revenue of $18 Million. The Latch merger makes it very clear that the public markets are betting on an industry #### **On the first day of trading as a publicly listed company, Latch saw a 4% increase in its stock. Latch’s merger to go public was one of the early announcements in the flurry of Proptech listings through a Special Purpose Acquisition Company. Latch’s beginnings as a smart home solution to rental apartments make it especially interesting to the Multifamily Owners who are looking towards technology to modernize their offerings but also understand their own position in a market which starting to bestow heavy multiples to proptech.** When Latch announced its merger with the Tishman Speyer vehicle, it was offered an **Enterprise Value of $1.05 Billion** on reported Net Revenue of $18 Million. The EV **multiple of 58.5x** is higher than any current publicly trading comparable. The deal also represents a transaction that is underwritten and executed by an industry incumbent like Tishman Speyer to create value from new startups in Property Technology. They are not alone. Lennar which is the largest homebuilder in the US has invested over $300 Million in startups which goes beyond the typical corporate innovation checks. Most of Latch’s revenue comes from booked revenues which capture 6-year contract terms in the pipeline but not currently represented in cash flow. It also showed a growth rate of around 50% for 2020 with losses of $66 Million. Latch’s product suite includes multiple modules but its resident access software remains its most popular product. Its growth prospectus, thus, is based on future successes of an API-driven operating system and additional focus on software integrating with multiple aspects of the asset and its residents. The Latch merger makes it very clear that the public markets are betting on industry instead of a product. There is an expectation of double-digit growth in the market of technology for apartment and rental living operations. The right platform approach to technology-driven operations for multifamily operations can help reduce operating expenses and improve flexibility to gain operating income. As Latch prepared to ring the NASDAQ bell, it also announced a variety of products for other verticals of the property market including offices and apartment upgrades. This is indicative that Latch’s success on the public markets is a test of the market it serves and its ability to translate to the whole asset-based industry. As a multifamily owner or operator, Latch’s performance on the public markets will be an indication of the appetite for change within the industry. As they grow to new segments, the percentage of revenues from these other segments will be an important metric to track the recovery and tech adoption from other verticals of the industry. ## When Industry Giants Fail: The Story of Countrywide URL: https://www.travtus.com/resources/when-industry-giants-fail-the-story-of-countrywide Date: 2022-10-30 Summary: Countrywide made the same decisions that most banks made with regards to technology. They decided to replicate and spend millions on digital assets without understanding the true potential of technology It has been exactly a decade since Lehman Brothers shuttered its doors as the first domino of the financial crisis. It was the event that changed the perception of the banking industry forever. It was also the case study that will be repeated over time about the lessons of liquidity and solvency. Lehman’s collapse was marked by its inability to find anyone willing to lend it any money. It could no longer get short-term loans, let alone raise balance sheet debt. The other underlying issue was the declining value of assets. The housing market’s depression resulted in a value base that very quickly changed all the metrics that define how solvent a company and its assets really are. The rest is history. But what is also great learning from this history is the impact it had on the future. The financial crisis would not have been a crisis if it was about the health of a single commercial enterprise or the mismanagement of specific companies. It was a systemic problem where the weak went down first and consolidation resulted in a handful of survivors existing in new structures of lower risk and greater regulation. > When the giants of an industry start to fail, it should trigger concern. It requires a real analysis of the what and the why along with the understanding of the impact on the market that it serves. We are at such an event with the collapse of Countrywide in the UK. Countrywide is the United Kingdom’s largest estate agency. It manages 125,000 properties around the country and owns 65 Agency brands all acquired by aggregating 850 agencies. In 2007, Countrywide was taken private by a consortium of Private Equity investors and relisted in 2013. During its IPO, it was listed at a £750 Million market cap. By 2014, it hit its peak at a market cap of around £1.4 Billion. This was followed by a period of underperformance and in January 2018 Countrywide announced an earnings warning. By August of 2018 it was unable to raise new debt with a failed bond offering. As I write this article, the Board has made an announcement for an Emergency Cash Call which is to be approved by August 28’ 2018. The market cap of the largest estate agency in the UK is languishing at £34 Million. It needs to raise another £140 Million for its “turnaround” which we should fairly call its “survival”. _So, what went wrong?_ A lot is being written about the competition from Online Letting Agencies and its impact on the traditional business models but we need to look at Countrywide intrinsically and the industry itself before we start blaming external factors. There is genuine truth in the fact that online letting agencies have a higher growth rate than the traditional incumbents but that is natural for any startup. When you start from scratch, growth rates will be higher than an incumbent. Also, less than 10% of the market is held by online letting agents and 90% is still with traditional incumbents. The problems that Countrywide faces are not driven by competition. It is more about the way that properties are managed and the industry itself. It comes back to liquidity & solvency. The problem is systemic very much like the problems with Lehman Brothers and it is only the first of the victims of a much larger tsunami of change. ## **We are in the People Business** A quick glance at the publicly reported financials of Countrywide shows the very base of the problem with the Agency business. It shows a Company whose Asset value lies in Intangible Assets and Goodwill. There is a lack of real assets including cash. Service businesses like Brokerage, Property Management, or Surveying are all based on intellectual property. The 2017 annual report of Countrywide acknowledges this with the statement _“Industry expertise in all areas of our business is key”._ Moreover, the acquisition of multiple brands has inflated the goodwill on the books with the brand value of all the 65 brands under the Countrywide umbrella representing large justifications of its Asset value. The “turnaround” plan for Countrywide included _“going back to the basics and bring back people who were with us before”_. This strategy ignores a very crucial problem. The business is not scalable. And to be the largest agency in the UK, you need more real “Assets” than disaggregated brands and a limited resource of people. Countrywide currently manages between 25-34 properties per employee on the payroll. This metric is actually quite standard for most traditional agencies. Based on publicly available information, the hybrid digital property managers are able to make an improvement of 10-20% on this metric and can manage around 35- 40 properties per employee on the payroll. This metric includes all the functions in a company like accounting, admin, agents, and management. In a self-confessed note in the Annual report, the management at Countrywide says that _“Centralization leads us to add substantial overheads to the group”_. When an enterprise scales, not being prepared for the scale is the biggest challenge that it faces. The Countrywide strategy to acquire brands and agencies was the perfect solution to show top-line growth and scale but it had no strategy or secret sauce for managing this scale. The economies of scale that are held as the truth of all management theory break down when there is no integration plan for the rapid growth. As Countrywide acquired more agencies, it hired more people to manage these acquisitions. For a company that has most of its use of funds being allocated to working capital and people costs instead of any real investment into an asset, it is no surprise that the sparkle of the top-line growth eventually turned on its profitability. Today, during a time of crisis, it will now have to face the challenges of redundancies and lease contracts on closed storefronts. But what is alarming for us as an industry is that these problems are not an issue just with Countrywide but a trend which is intrinsic to the ways properties are let and managed today. “We are in a people business” is a trademark for most agency businesses. However, the industry is also in the real estate business. The landlords who are paying for tenancy fees, management fees, and sometimes also ground management fees are driven by yields. If there is any way to help improve their yields, then the industry must focus on that, or else it should come as no surprise when it collapses under the weight of its own overextension. ## \*\*\*\* ## **We can (not) do Digital** The 2014 Annual Report from Countrywide uses the word “innovation” 5 times. It announces new channels in addition to the storefronts like digital portals and applications. More people are hired to manage these new channels. Over £2 Million is spent on marketing redesign and online presence. Large IT teams and partnerships are set for “transforming technology infrastructure” with the creation of a common platform and integrated VOIP.  Countrywide announces the launch of “Traveltime” and “ Launchpad”, its digital solutions for property search and landlord tools. Countrywide also launched a digital fixed fee proposition to customers which was a reduced fee offering without full service. Today, in 2018 the same Annual report mentions _“ We need to define what digital means to us as an organization_”. Like most in the industry, Countrywide jumped onto the bandwagon of technology solutions as a trend follower. Their digital platform competed with their full-service proposition creating confusion for both the employees and the customer on the expectations of the actual value proposition. This confusion is not just a problem for Countrywide. It is a systemic problem for all traditional and new online property managers. While full-service agencies are ramping back on services, the hybrid platforms are adding new full-service options. Countrywide made the same decisions that most banks made with regard to technology. They decided to replicate and spend millions on digital assets without recognizing what the true potential of technology really is. Its priority in 2018 as mentioned in the latest annual report continues to be to improve its legacy IT and contact center which hinder all scalable growth and in this case, the efficiency of scale. ## **The Day 1 Philosophy** Startup founders often worry about the consequences of someone “copying your idea”. It is a very irrelevant fear. There is no reason why any business would not have evolved to an improved vision of itself by the time the “copying” happens. The growth of Amazon from an online bookseller to the master of the universe is based on the Jeff Bezos trademark philosophy of _“Day 1”_. It did not matter how many e-commerce businesses “copied” or incumbents created new channels, it did not stand still for a single day, building a scalable enterprise. The Day 1 philosophy strives to keep a startup mentality at every stage of the scaling of a business. It guards against the contentment from success. Day 2 is when irrelevance begins. It is followed by pain, decline, and death and it can be slow but it is inevitable. The property management industry needs to find its roots back to “Day 1”. The industry has been using the shield of competition and macro-economic trends to ignore that it is languishing in Day 2. The strategy to scale in property management & agency has not yet been discovered. The traditional incumbents are using their financial lead to acquire a winner’s position. The new entrants are using technology to expose changes in the demands in the market and also get to a lead but both will eventually face the same issue of managing scale. Tomorrow if Countrywide was acquired by another incumbent or online player, the challenges will remain unchanged. The customer will continue to demand better yields and a real solution will need to solve for scale. Till then, the coming years will continue to see more businesses fail due to a systemic industry problem. ## Centralize your operations with (Artificial) Intelligence URL: https://www.travtus.com/resources/centralize-your-multifamily-operations-with-artificial-intelligence Date: 2019-08-14 Summary: A large NMHC 50 operator needs to look at prioritizing long-term initiatives which are not embedded in software but in data. The biggest challenge faced by the owners of 2 Million apartment units in America right now is – how to centralize operations > **The centralized services model for multifamily consolidates functions like leasing, support and dispatch into shared teams that serve many properties at once, instead of duplicating them site by site. Done with AI and your own data, a central team handles far more volume with consistent quality — you use the advantage of scale to run operations without feet on the ground.** The NHMC Top 50 owners account for 10% of the US apartment stock. The percentage of institutional ownership of the market increases every year. However, most aggregated portfolios are still managed as individual assets which are traded as individual operating entities. The pressures of fund deployment and asset acquisition often result in increasing ownership but with a diminishing return for scale. Leading operators are increasingly looking to build a strategy that can leverage their scale and the blistering progress in Artificial Intelligence platforms like Adam, can now be capitalized on to utilize the benefits of scale and gain front-line intelligence. One of the main benefits of building an operating platform that is powered by Artificial Intelligence is the ability to centralize and also customize without the need for ‘feet on the ground’. But in order to achieve this, rental operators need to make some fundamental strategic shifts. ### **Vision first, requirements second** > ‘Leaders also often think too narrowly about AI requirements. “ > > _Building the AI-Powered Organization, Harvard Business Review_ For most real estate operations teams, the idea of a centralized operations team seems ridiculous. To them, it takes away their insight from being on the ground and yet, a similar approach has been working for many single-family rental institutional operators for some time. To bring the real change you have to think differently. The biggest challenge an organization needs to overcome when trying to centralize its operations is to stop thinking about the existing organization. Most operators will look at centralizing certain job functions like marketing, leasing, and possibly service dispatch. They use their understanding of existing jobs people occupy nad them try to centralize them. However, they are unable to imagine solutions for tasks or problems they are unaware of. Very little time and budget are given to defining the full scale of the problem. In order to successfully centralize operations and operate at scale, the business needs to use the benefit that their scale provides – DATA! However, despite having the scale most organizations don’t have useful data to leverage and learn from. When formulating your AI Strategy, the worst mistake any operator can do is to apply a _‘bid based_’ requirement created by the arbitrary demands of the business. First, you need to understand the full scale of your operations, understand the hidden facets of your business that you don’t know to exist. An effective AI strategy will always start with capturing the right data. By doing so you will unearth the tasks you never knew your teams were performing and will discover both opportunities for centralization and problems that could derail the change. If you try to define requirements prior to data collection can result in the worst-case scenario where some jobs get centralized and others stay localized creating communication barriers, the unclear path of responsibility, and general corporate politics. ### **Institutional learning is a strategy** > “At most firms that aren’t born digital, mindsets run counter to those needed for AI.” > > _Building the AI-Powered Organization, Harvard Business Review_ The best operations teams are often led by experience-driven leaders who build a hierarchy for escalation for themselves. Starting the centralization strategy with the goal of collecting front-line intelligence allows for the human chain of command to be broken. Employees can now augment their judgment by using field intelligence to more transparent problems. Regional hubs can now become a place for problem-solving and decisions instead of fire-fighting. These decision teams can in fact be located anywhere and work iteratively on an ever learning and evolving operating platform. In order to make the most of institutional learning, the teams working with artificial intelligence should have a mix of skills and perspectives. It is important that operations, tech, marketing, and asset teams work together to address organizational learning and solutions in order to get the most value of the business’s scale. ### **Operations mean content and data** > “Without data you’re just another person with an opinion.” > > W. Edwards Deming Property operators mostly consider the role of their operations teams to be reactionary. They are either working through the unexpected or following up on the ongoing. There is a precondition to describing the job as one of ‘legs on the ground’. While, there is no denying the importance of the field visits, follow-ups and inspections, there is a gaping hole in the industry towards the role of operations in content and data accumulation. Centralizing operations teams can help remove some of the bias of opinion which interferes with scientific decision-making. Redefining the job description of operations to include data collection helps your scaling business to look at problem-solving as their job. The greatest impact of data collection and institutional learning is on content creation. Most operators today have unclear content strategies. It is often separated into marketing or community communication. Corporate communication teams prepare the content based on legal and marketing feedback but are often devoid of the detail that the field intelligence captures. Data does not make your teams smarter but it provides insight. The insight can identify a pain point and help your teams prevent it rather than reacting to it. The appetite for change that needs a balance of feasibility, time and cost is found mostly in businesses that are looking to scale and build defendable IP for their industry. A large NMHC 50 operator needs to look at prioritizing long-term initiatives which are not embedded in software but in data. The scaling strategy for most rental operators is currently focused only on acquisition. It is glamorous and is described in millions of dollars invested. However, the challenge of operating these assets as a single brand is not a small problem to solve. ### **The biggest challenge faced by the owners of 2 Million apartment units in America right now is – how to centralize operations. In order to answer these questions, they need a platform based on data and learning.** ## How to hyper-automate your residential operations URL: https://www.travtus.com/resources/how-to-hyper-automate-your-residential-operations-and-centralize Date: 2019-08-14 Summary: Gartner predicts Hyper-automation will be the next big trend in 2022. Here is how residential operations can pursue hyper-automation to automate end-to-end operational streams beyond software. One of the top trends on the Gartner Top Strategic Technology Trends for 2022 is [Hyperautomation](https://www.gartner.com/en/information-technology/insights/top-technology-trends/top-technology-trends-ebook). The Gartner Glossary defines Hyperautomation as: > "... a business-driven, disciplined approach that organizations use to identify rapidly, vet, and automate as many business processes as possible. Hyperautomation involves the orchestrated use of multiple technologies, tools, or platforms, including Artificial intelligence (AI), Machine learning, Event-driven software architecture, Robotic process automation (RPA), business process management (BPM), and intelligent business process management suites (iBPMS), Integration platform as a service (iPaaS), Low-code/no-code tools, Other types of decision, process and task automation tools." ### **How is Hyperautomation relevant to residential operations?** **1 in 3 residential operators has the mandate to centralize and automate their operations by 2024.** The pressures of labor supply, customer expectations, and data expectations of LPs have prioritized the need for automation initiatives. But the domain of onsite management is very varied and opaque, which means that simple automation has a meager impact on productivity or bottom lines. But, the pursuit of hyper-automation can automate end-to-end operational streams beyond standard rules by combining a domain-trained and sophisticated machine learning platform with conversational intelligence. ADAM is the perfect partner in the industry's journey to hyper-automation. ### **Getting Started for Success - Process Mining** Organizations like to "automate" with bias. The department with the most budget, churn, or other indicator defines a need, and automation proposals are submitted. However, this method does not apply to sustainable hyper-automation. The first and most crucial step for transformation is Process Mining. In simple words, it's the activity of understanding the onsite tasks and gathering operational intelligence that can then help alter the existing process before automating. Traditionally, a group of internal or external consultants would 'shadow' or simulate with your onsite teams to map their tasks and then suggest improvements. However, the biggest problem that large operators face is that every community is unique. It may be a consequence of inherited practices from a previous owner or just the evolution for customization for a single building. Also, a lot of operational peaks are seasonal. The greatest challenge in operations is exception handling, which requires a longer time frame of observation. Operators need to study their teams effectively over an extended period, multiple channels, and communities to process mine effectively. However, using a conversational interface with domain training like ADAM, this data is gathered accurately within weeks. ADAM behaves like a silent observer within the onsite teams and helps cluster trends and identify the patterns of operational bottlenecks. ### **Preparation for Automation - Digital Operations** Technology companies often talk about technical debt, but real estate operators rarely discuss the operational debt. When applied in an onsite scenario, processes designed in centralized offices are often not fit for purpose. On the contrary, complete flexibility and freedom for onsite operations are double-edged swords. It allows an excellent manager to perform to their potential but creates an intelligence void and loss of transparency for the central organization. > When designing your 'digital operations, ' adaptability is crucial. It needs to provide the flexibility of applying consistent, centralized, and often 'branded' processes to make sense for scale and local variation. Onsite Digital Ops requires an event-based automation architecture instead of simple workflow software orchestration. ADAM provides this. Any process or automation can be triggered, either from centralized policies or replaced by local tools which are more relevant to a community. ### **Total Experience Management - Conversational UX** Any automation project needs to include both the employee and customer experience as a single strategy. Technology should not create friction for change. Most centralization and automation strategies are underpinned by the deployment of mobile apps or software for users to engage with. Staff training is an important factor and the strain of 'learning tech' is a pain point for property teams. After deploying a solution, switching becomes impossible because of the pressures of change. No one wants to communicate new app downloads to residents or retrain staff on new systems.Hyper-automation strategies mandate flexibility in architecture that can only be achieved with a natural and intuitive UX. Conversational UX provides this, with no learning curves or downloads allowing frictionless change. When ADAM has a conversation with a resident, for example, to update occupants’ contact details, he can seamlessly decide to trigger multiple RPA processes which can be swapped at any time without any experience of change to the user. Employee and customer confidence is high when their experience is seamless. Natural language is the best fit for a flexible interface but it requires a quality of interaction that can only be achieved by dedicated domain-based conversational engines that straddle all channels and the entire domain. Deploying a Conversational UX that doesn’t understand the breadth and nuances of the entire Real Estate domain will lower confidence. ADAM has created this trust with over 80,000 residents who treat him like a human member of their community teams to achieve automation while retaining the personal touch. ## Printing & Connectivity : The hidden service URL: https://www.travtus.com/resources/what-do-renters-want-from-internet-and-what-do-landlords-do Date: 2019-08-14 Summary: Travtus and Wiredscore teams combine conversational intelligence and best practices, to highlight the need and operational impact from connectivity support demanded by residents. A WHITE PAPER BY [TRAVTUS](https://www.travtus.com/) & [WIREDSCORE](https://wiredscore.com/) , WITH INSIGHTS FROM CONVERSATIONAL INTELLIGENCE AND CERTIFICATION **30% of multifamily renters will use an onsite printer?** Everyone needs the internet. There is really no need to write a white-paper about it. However, the question of what do renter’s want, is still a fluid question and that answer changes every leasing cycle. The need for data to understand your resident experience is crucial but what’s more important is to take action on these insights. It is with this goal, that the Travtus and Wiredscore have combined conversational intelligence and best practices, to highlight the need and operational impact from connectivity support demanded by residents. ### **Why is connectivity a multifamily service?** Residents are working from home more, and multifamily owners and operators are now their landlords for both home and office. As a result, service expectations from apartment communities are increasing, and one of the most trending amenity today is digital connectivity and believe it of not, printing. Adam’s analysis of 50,000 conversations and 100,000 emails from customers of Multifamily residential assets across America, show that **“I want to print & copy**” in within the **top 5 customer goals**. In some communities, it is more than the number of maintenance requests being submitted. Residents are often sending unsecured documents to community inboxes and requesting the onsite office to print them on their behalf. These documents can range from powerpoint presentations, to visa documents and even return labels for online shopping. These “hidden” services are a reality for onsite teams . They are already providing the service , layered within the [long tail of interactions](https://www.travtus.com/insight/are-multifamily-properties-understaffed) with their residents. However, the significance and cadence of these requests is lost in the chatter of conversational operations. Adam’s conversational intelligence tracks renter interest in connectivity and business services fluctuating with work from home announcements. The forward trend analysis confirms that that residents chatter about connectivity remains higher than 2019 levels. During the pandemic, a direct correlation was seen between lockdowns and the trends in connectivity related chatter. The first lockdown in the summer of 2020, had a wave of chatter about public health and safety. But, the nature of chatter was very different in the summer of 2021, when residents faced the second wave of lockdowns. Adam saw a 300% rise in conversations about connectivity topics. These conversations start with a prospect asking a common question - **Who pays for the internet?** and continued through the resident journey with managing their internet connections. ### **How do we provide good and secure service?** The Wiredscore Home certification has identified five key areas to ensure that Multifamily communities can deliver a quality connectivity experience to their residents. ![](/assets/blog/what-do-renters-want-from-internet-and-what-do-landlords-do__Multifamily_resident_connectivity_chatter_by_outcome.png) 1. **Stability** \*\*The internet you provide residents must be stable and capable of providing internet networks that not only perform but can also survive disruption of service. The hidden cost of connectivity support is often lost in vendor negotiations. Adam’s chatter shows 1.5x engagement when the wifi of a community is not working. 2. **Speed** \*\*The needs for each resident are different and it is important to offer a range of connectivity packages which can fit their needs and pocket. 3. **Security** _\-_ Your residents are using their homes like offices. All data shared on your networks must be secured. Wifi offerings should always be safe and secure via a protected network. Printing exchanges must be set up to prevent data breaches from inboxes. 4. **Setup**\- When a resident moves into your community, the most popular connectivity related question is - How do I set up my internet? If you chose to provide network connectivity or not, Adam notes that most residents request support with going online within the first three days of their move. To reduce the “chatter support” from new move ins, having a range of internet services already set up and ready to plug in makes it easier on both onsite employees as well as residents. 5. **Services** _\-_ Residents have expectation of service if connectivity is available. This can include assistance with getting online or public wifi access. Adam’s analysis sees an increase in support needed with regards to cable and internet hardware return. The onsite teams, may consider managing connectivity as outside their scope of service, but like most items in operations, there is scope creep with regards to co-ordinating resident’s connectivity needs. Business center facilities are now also becoming an essential amenity which needs to be managed with the same excellence as other aspects of the resident’s experience. The most crucial aspect of providing good service is to maintain a culture of “community listening”. Tools like Adam, who assist with automation as well as information, are crucial to stay on top of the latest needs of residents and also react to them with speed. When looking at the mapping from Wiredscore as well as Adam’s field intelligence, we see a reinforcement of the message from multifamily renters. ### **How much do residents really care about this stuff?** We are in one of the highest inflationary markets and rents are being priced up without any need for carrots like amenities. Most operators, will ask, if there is any financial upside to adding more service to the offering. “Will it rent me more apartments”? ![](/assets/blog/what-do-renters-want-from-internet-and-what-do-landlords-do__Trend_of_Digital_Connectivity_Chatter.png) Our joint analysis of Adam’s conversational chatter with Wiredscore’s certification performance, tells us the the question is really about optimising a service that we are already providing. On site teams are already providing support for business services informally. They are also co-ordinating between your residents and internet companies. The calls about internet outage at the community center, comes first to your resident services desk. The question that we ask about value is no longer - is this a service we should provide? , because unknown to most operators, they are already connectivity service providers. There is however, an opportunity to do it well. Using a conversational platform like Adam, allows operators to respond to support issues and new requests without additional burden to the onsite teams. Queues of tickets can be dispatched directly to the service provider, instead of creating a triangle of support between the office, resident and the vendor. Similarly, a Wiredscore certification allows for a service propositioned to be designed for best practices such that it adds a mark of quality but also reduces support that can blimp the resident experience. Adam maps every resident’s journey. He notes everytime your customer contacts your teams. We have seen that 30% of all residents have at some point reached out about printing services and 10% have had some interaction about internet connectivity. These are not numbers to ignore. A good resident journey, needs tools to address these needs. ## Automate resident support using machine learning and email URL: https://www.travtus.com/resources/automate-resident-support-using-machine-learning-and-email Date: 2019-03-11 Summary: On average, the industry gets 35 emails per 100 units every month from customers needing support. The right machine learning platform can help solve this problem. Multifamily operators use many channels for offering support to their residents. But the most popular medium for customer service management continues to be email. > **On average, the industry seed 35 emails per 100 units every month from customers needing support.** As a result, your property inboxes are flooded with resident requests, vendor queries, automated notifications, and internal memos. Email management is now an essential skill that can help optimize property operations. It is the most common mode of collaboration and also a record of information. Unfortunately, property teams often use the long chains of email as both a mode of communication and a repository of knowledge. As a result, it is becoming more difficult to find actionable intelligence from our inboxes. There is a solution to finding actionable insight from the clutter of community inboxes. Sophisticated industry-specific language models can exact performance and trends to help reduce workload and also automate processes. It can help identify the emails that need attention and add them to queues or tasks lists by both priority and department. Our email data is an excellent source of big data, which helps drive a better understanding of our customers and properties. According to the McKinsey Global Institute, professionals spend 28% of their time on emails. The onsite teams at our properties are often dealing with email overload. > **Our teams can be more productive by working through automated task lists generated by the models analyzing emails.** This approach gives them the time and resources to respond with a resolution without distraction from their immediate tasks. It also helps move the priority of a request by tracking the time commitment for a job that is at risk of being lost. Ultimately, the efficiency of your team helps improve both resident experience and retention. Owners can share an effective team working through tickets over multiple properties. We can now reduce the cost of support while increasing the long-term value of each resident, without any need for change management by using the humble email inbox with some cutting-edge machine learning.