We have no survey of which AI tools real estate funds use, so this page names none. What we can say is what funds use AI for: four jobs, each with a different test. Titleman covers all four, and the measured results below are what to hold any tool to.
Prepared by Titleman AI, on behalf of the Titleman team. Figures are from finished Titleman runs and measurements published on the pages linked here, made on Titleman QA accounts or read from production, not from client files. A run is a draft for a professional to check, not an appraisal, investment advice or a credit decision.
The four jobs
| Job | What a fund asks | What we measured |
|---|---|---|
| Underwrite an acquisition | Does the deal clear our return at this price? | Median about 58 minutes (44 runs, 30 days to 18 October 2026) |
| Screen many deals or sites | Which of these deserve an hour? | Three sites in 48.7 minutes in one run; no 20-site batch measured |
| Check a deal package | Do the documents agree with each other? | One six-file package: 1,788 items checked, 0 left unchecked, 6 red flags, 46 to review |
| Source under pressure | Which owners and loans will transact? | 597 of 2,879 multifamily buildings of 5 or more units carried a distress record in one run |
Underwriting, on one run
The run was an internal QA acquisition underwrite of a roughly 100-unit building, asked the way a fund would ask it. It took 81 minutes and returned 147 figures, each tied to a record. It stated the refinance case: coverage of 1.53x as reported fell to 0.97x at 6.5% interest-only and 0.79x at 7% on a 30-year schedule. It also said what it could not settle. With no rent roll or T-12, the USD 442,327 a year between in-place and market NOI was unproven. See how a fund evaluates AI underwriting tools.
Deal review
A deal review reads the files you send and marks each figure verified, to review or red flag on a copy, so your originals stay as they were. Code recomputes formulas and settles figures stated in more than one place; the model checks the rest against the market. Every review opens with a count, and a short, specific list is what a person who signs can act on.
Sourcing
Starting from every building in scope, not from listings, finds owners who have not decided to sell. The run above said what it could not read: 1,611 of the 2,879 buildings had no tax balance read, so they were unchecked, not clean.
What it does not replace
AI does not replace the investment committee's judgment, a licensed appraisal, a title search, a physical inspection or legal review. It does not state a bid price.
To see which job fits first, read AI tools for real estate or send a deal you already underwrote and compare.