Ocrolus

Ocrolus is a document AI company founded in 2016 and headquartered in New York that converts financial documents into structured, decision-ready data for lenders. It targets the manual bottleneck in credit operations: bank statements, pay stubs, tax forms and identity documents that arrive as scans, photos or PDFs and must be read, validated and checked for tampering before an underwriting decision.

The platform pairs machine learning with a review layer to classify documents and extract fields at accuracy the company reports above 99 percent, then adds analytics on top: cash flow analysis for small business funding, income calculation for mortgage, and fraud signals such as altered files or inconsistent metadata. Ocrolus states it processes about 750,000 credit applications a month for more than 400 lenders, among them Better, Enova, SoFi and Fora Financial. In March 2026 it released automated conditioning for mortgage lenders, generating and tracking underwriting conditions against selling guide requirements inside Encompass.

1. Document Understanding: classifies and extracts fields from bank statements, pay stubs, tax forms and IDs at reported 99%+ accuracy.

2. Human-in-the-loop review: model output is verified before release so the data can support regulated credit decisions.

3. Cash flow analytics: derives deposits, revenue, NSF events and balance trends from statements for small business underwriting.

4. Fraud signals: flags altered documents, inconsistent metadata and values that disagree across a borrower's file.

5. Automated conditioning: generates underwriting conditions from selling guide rules, matches documents to them and tracks status, generally available 1 April 2026.

6. Scale: about 750,000 credit applications a month for 400+ lenders, with Encompass integration for mortgage teams.