How does an AI underwriting analyst learn a firm's style?

An AI underwriting analyst learns a firm's style from the firm's own work: its Excel template, its past models and memos, its buy box and the corrections the team makes on each deal. The more deals the team reviews with it, the closer the output gets to what the firm would have produced itself. What it learns from - Your template: tabs, line items, formulas and formatting, so the output drops into your process. - Your assumptions: rent growth, expense growth, exit cap rate spreads, reserves, hold period. - Your buy box: markets, property types and the return thresholds that decide a pass. - Your edits: each time an analyst changes a figure or a sentence, that is a signal for the next deal. How to set it up in the first weeks 1. Send three to five past deals with your final models and memos. 2. Run two or three live deals side by side with your analyst and compare. 3. Correct what is off and say why: "we never underwrite rent growth above 3 percent in year one". 4. Save the rules that come out of it as written assumptions the AI applies every time. What it should not learn It should not learn to hide uncertainty to match a house view. When the market does not support an assumption, it should say so, even if the template expects it. What to measure Time from package to first draft, how many figures your team changes, and whether those changes shrink deal by deal. Titleman works as an AI analyst that writes into your template and keeps your rules from deal to deal: the more deals you review together, the closer it fits your team. See how to start: https://titleman.ai/hire