LIHTC tenant file review
Upload a certification packet. Get citation-backed findings with page-anchored evidence — income, assets, rent, student status, and the layered-program rules that trip up even careful files.
Now onboarding a small number of pilot partners.
A single certification packet or a whole batch — TICs, verifications, statements, leases, the works. Scans and phone photos included.
Full Social Security and account numbers are destroyed before any AI reads a page. Then every document is cross-checked against every other one — and against the rules in force on the certification's effective date.
Each finding carries its severity, the governing citation, the exact page the evidence sits on, and what to do about it. When the file doesn't support a conclusion, it says so instead of guessing.
The engine is tested against adversarial files with planted defects, traps, and clean controls. A finding has to earn its way out — a wrong accusation against a tenant's file is the one failure we refuse to ship.
Findings cite IRC §42, the Treasury regulations, HUD Handbook 4350.3, the 8823 Guide, and state HFA rules — with effective-date awareness, so a pre-HOTMA certification is judged by pre-HOTMA rules.
LIHTC rarely travels alone. HOME, HTF, PBRA, PBV, and HCV layers each bring their own limits — the engine reconciles them to the most restrictive rule instead of testing the wrong ceiling.
For property management companies, owners, syndicators, and housing finance agencies. Findings and analyses are AI-generated review assistance for housing professionals — a second set of eyes, not a replacement for your reviewer and not a compliance certification.