Voyager Suite
Breeze
Virtuoso
Investment Suite
RentCafe
CommercialCafe Suite
Elevate Suite
Procure to Pay Suite 
Energy Suite
Aspire
RightSource
Care Suite
Matrix
WorkCafe
Yardi Kube Suite
ScreeningWorks Pro 
Yardi Corom
Data readiness for AI in real estate means knowing what data you have, where it lives and whether a model can reach it. That last part is the hard one. Decades of accounting, leasing and maintenance records sit across systems that were never designed to talk to a language model.
Episode 6 of The Capital Stack takes on that problem. Host Jace Swank, senior director of investment management solutions at Yardi, sits down with Shivani Kumar, who leads Yardi’s Applied AI team, to discuss what data readiness requires, who needs to be part of the decision process and where investment managers are seeing returns.
Key takeaway
Investment managers who treat governance as an ongoing function, and who bring operations, technology and executive leadership into the same room, move faster than those waiting for their data to be clean enough to start.
Data readiness is an aspiration, not a destination
Kumar is clear about the limits: data readiness is an aspiration, not a goal you achieve. Gaps will persist and cleanup will be an ongoing effort. Focus on these four things:
- Know what you have: What information exists, where it lives and which systems read from and write to it.
- Put governance in place: Larger organizations often run a formal master data management function. Smaller teams need at minimum an index of what exists, a defined process for changes and traceability.
- Tie it back to performance: The data that matters most is the data connected to the net operating income.
- Expect calibration: AI consumes information differently depending on the use case, so pulling the right data for a specific workflow takes tuning.
The calibration point is where most first attempts stall. A common instinct is to pair an existing data set with a data dictionary and expect the model to interpret the schema on its own. In practice, this approach has limits, and the work often requires a purpose-built interface between the model and the data.
The three-way partnership behind data governance
Effective data governance depends on a three-way partnership, not a chain of requests:
- The COO understands the day-to-day decisions and the information they require, and needs patience, because aggregating what sounds simple often takes several steps.
- The CEO articulates future demand, shaping what the data must support months from now rather than today.
- The CTO or CIO owns the answer. Data governance stems from and must be managed within the technology organization.
That requires a shift. Technology was once a service organization you filed requests with. It now has to be a partner in the room when objectives are set, because it owns the stack.
“It needs to be seen as a three-way partnership, not as supply and demand,” said Kumar.
Where AI delivers value in real estate
AI can reason, read documents and produce language, which widens what an organization can reasonably delegate to a system. Kumar identifies three areas where the return shows up:
- Analytics at your fingertips: Connecting an LLM to portfolio financials allows real-time investigation. A question about why an asset is underperforming its market resolves in minutes rather than months.
- Operational efficiency: The scope of what a business can trust a system to handle is far wider, including complex decisions with an audit trail. Budget variance analysis is the clearest example, with a model running the analysis and drafting the notes end to end.
- Revenue growth: Faster residential turn times support occupancy, and pricing can now draw on walkability scores, school rankings and ZIP code data in one place.
A fourth benefit resists measurement: better experiences for residents, tenants and the people doing the work. It carries no direct NOI impact, but it shapes whether the other three hold up over time.
What AI adoption requires
Organizations fall into two groups. The first has a strong appetite for the technology itself and will test any application that might add value, including tasks that were previously too dependent on human judgment to automate.
The second group is larger and treats AI as one way to reach its goals. Adoption there resembles any structured change program. It is slower but more durable, and in Kumar’s view it produces the bigger impact, because those organizations assess their current state honestly.
“AI by itself adds value, but AI alongside process change is transformational. AI can change how people work, only if people are willing to change how they work,” said Kumar.
Building on a data foundation
What AI can do with your data depends on how well that data was structured to begin with. Yardi has spent more than 40 years working out how real estate data should be captured. Yardi Virtuoso is the intelligence built into that foundation, available across the Yardi products firms already use to run their real estate business.
On the investment side, connected data is what makes AI useful. Yardi Virtuoso for Investment Management covers the entire investment lifecycle, from acquisition through to investor reporting, on a single connected data set. Because it draws from the system where your whole portfolio lives, its answers reflect everything you manage. Beyond answering questions, it can also handle routine work, such as creating a deal from an offering memorandum.
Watch the episode for the full conversation on getting data ready for AI.
