# Fync AI

> Fync AI helps CRE lenders turn property packages into cited underwriting data, review queues, and credit memo drafts.

## What it is

Fync AI is a CRE (commercial real estate) underwriting intelligence workspace. Companies define their underwriting standard — fields, documents, required checks, synonyms, output expectations — and every property workspace then applies that standard to the package it receives. Every fact is cited back to its source document, public data lookup, or vendor report.

## How it works

1. Define standards: workspace field dictionary (labels, keys, types, units, synonyms), required documents, output templates.
2. Create the property workspace.
3. Upload the borrower package (PDF, 50 MB limit, 39 supported document types).
4. Extract cited facts — Fync AI classifies each document, OCRs and parses pages and tables, extracts to your field dictionary, and verifies every value against its source page.
5. Enrich with public + vendor data — 19 enrichment adapters covering Census, FEMA, BLS, permits, assessments, HUD FMR, FRED, CoStar, title, Phase I ESA, PCA.
6. Detect gaps and conflicts — surface missing critical fields, conflicting numbers, unverified citations, and stale data.
7. Review exceptions through 14 review queues — reviewers see what needs attention, open evidence, choose a resolution, and log a reason into the audit trail.
8. Ask cited chat — structured-data-first answers with document-evidence fallback; every answer cited.
9. Generate credit memo drafts — dependency-gated, from approved citation-verified facts.

## Citation and approval model

Citation verification (machine) is separate from reviewer approval (human). A fact can be extracted and source-matched without being approved. Reviewers approve, reject, edit, resolve conflicts, and write decision reasons. Every action is captured in the audit trail with who, when, prior value, and reason.

## Platform

Persisted artifacts. Typed workflows. Workspace-level isolation. Encryption in transit and at rest. Role-based access. SOC 2-aligned practices. Full audit logging. The developer view exposes pipeline state — artifacts, runs, vector chunks, env presence booleans — without leaking secret values.

## Numbers today

- 18 Knowledge Base sections
- 14 Review queues
- 19 Enrichment adapters
- 39 Document types
- 8 Generated output types
- 5 Package source categories

## Who it is for

CRE underwriting teams at lenders, equity investors, and credit committees that intake property packages, screen acquisitions, prepare for quote committee, replace manual fact checking, run public-data lookups, and produce credit memo drafts.

## Pages

- [Home](https://fync-ai-marketing.pages.dev/) — product overview, workflow, feature pillars, data network, use cases, platform, and contact.

## Machine-readable discovery

- [Sitemap](https://fync-ai-marketing.pages.dev/sitemap-index.xml)
- [Robots](https://fync-ai-marketing.pages.dev/robots.txt)

## Contact

- Book a demo: https://calendly.com/fync-ai/demo