Verification is where loan files stall. Employment gets confirmed by phone, income gets recalculated by hand, and assets get checked against statements that arrived by email three days ago. Verification Engine replaces that with source-validated data and AI document extraction, so the file is verified before an underwriter opens it.
Verification Engine pulls verified employment and income data directly from payroll, bank and asset networks in real time. Where a source connection is available, the data arrives already validated rather than assembled from documents after the fact.
Where documents are the only source, the platform extracts structured data from paystubs, W-2s, tax returns and bank statements using OCR and machine learning. The extracted values populate the loan file directly instead of being re-typed.
Every extracted value is compared across the documents in the file. Where a figure on a paystub disagrees with a tax return or a bank statement, the inconsistency is flagged for underwriter review rather than passed through silently.
The platform screens submitted documents for signs of tampering and for synthetic data anomalies, surfacing them at intake rather than at audit.
Verification Engine feeds the same loan file that Disclosure Hub and Compliance Monitor work from. A verified income figure is available to disclosure generation and to the compliance rule engine without being entered a second time.
Verification of employment, verification of income and verification of deposit — the three checks that confirm a borrower can support the loan. CliQloan automates all three from source data and submitted documents.
Paystubs, W-2s, tax returns and bank statements are supported today through OCR and machine learning extraction.
No. It removes the re-keying and the manual cross-checking, and surfaces inconsistencies for a human to decide on. The underwriting decision stays with the underwriter.
Submitted documents are screened for signs of alteration and for synthetic data anomalies at the point of intake.