Loan Underwriting

AI reviews borrower files against lending rules and flags underwriting issues before a formal decision.

Business impact

• Underwriting cycle time — The workflow shifts document checking earlier and reduces manual review time.
• Exception rate — One implementation focuses on surfacing exceptions so staff can review them sooner.
• Manual review workload — Standard tasks are automated so underwriters can focus on exceptions.

Data requirements

• Borrower documents (text) — Used to check file contents against underwriting requirements.
• Loan application data (tabular) — Supports qualification and eligibility review.
• Investor guidelines (text) — Defines the rules used to test product eligibility.
• Lending guidelines (text) — Sets the criteria for pre-underwriting review.

AI methods and techniques

• Document review — Reads borrower and mortgage documents to find relevant underwriting information.
• Classification — Decides whether file items meet product or guideline requirements.
• Exception flagging — Marks missing information or likely underwriting conditions for human review.
• Question answering — Answers loan-qualification questions for loan officers.

Industries

Real-world evidence

4 documented case studies on record.

Companies using this: Griffin Funding, Longbridge Financial, Newrez, Fannie Mae.