The AI that cites its work.
AOS-1 is the AI inside AppraisalOS. It sits beside the field you are editing — reading the contract, helping with comps, drafting narratives, flagging what doesn't line up. Answers carry source context when it is available. AI output may be incomplete or inaccurate and must be reviewed by the appraiser.
How it works, step by step.
- 01Parses the order on intake
Order details parsed on intake — lender, client, subject address, and due date pre-fill the assignment. Loan type, intended use, and effective date are confirmed by you during form completion.
- 02Pulls source documents
Assessor card, MLS sheet, flood cert, contract, prior appraisal — indexed and searchable.
- 03Proposes values with citations
Supported fields it touches can show a source you can click. Nothing is eyeballed.
- 04Drafts narratives with sources
Neighborhood, site, and comparable commentary — each claim tied to a citation chip.
- 05Flags what doesn't line up
AOS-1 QC flags inconsistencies for you to review, such as GLA vs sketch or sale price vs contract. Math and compliance are rules, not AI: calculation engines compute ANSI Z765 GLA and HPI time adjustments, and Compliance Check checks UAD 3.6 enums and conditional rules.
- 06Asks when it's unsure
Low-confidence fields surface as a question for you, not a silent guess.
Everything in this feature.
Supported answers draw on the documents and data sources attached to the order — not from open training data.
Click any AOS-1-proposed value to see the source document, page, or data point.
Inline guidance on supported UAD 3.6 fields and sections — the definition, the allowed enums, and a proposed value for you to accept or reject.
Neighborhood, site, and comparable paragraphs with in-line citations.
Flags inconsistencies across the URAR (GLA vs sketch, sale price vs engagement letter, etc.).
Metered realtime voice chat: ask out loud and hear the answer. Included in Founding Intelligence.
When it can't be sure, it asks. Silent hallucinations are a reviewer nightmare — it doesn't do them.
Cited work changes what reviewers see.
A reviewer opening a report grounded in clickable citations spends less time suspicious and more time confirming. That's the entire point — reduce friction on the way out the door, not add more fireworks to the way in.
Quick answers.
- No. AOS-1 is AI: software built into AppraisalOS, not a person, and it says so when asked. It assists; the appraiser reviews its output and makes every decision.
Ready to see it live?
30-day evaluation. Office $49, Pro $99 or Complete $199/month after the evaluation. No card to start.