Tidalwave AI tops Claude in mortgage underwriting accuracy

"Tidalwave's SOLO scored 95% accuracy in that category, compared with 42% for the baseline model. The biggest performance gap was in yes-or-no compliance checks the questions used to flag issues such as payroll mismatches, undisclosed debts and suspicious transactions."

"Diane Yu, co-founder and CEO of Tidalwave, said in an interview with HousingWire that the outperformance is intentional because SOLO strips out personally identifiable information (PII) before processing requests. This confirms the belief that generic large language models (LLMs) use PII data sent to them, which is a violation of customer privacy in mortgages."

"General-purpose models analyze loan files as text, while SOLO is integrated with underwriting systems used by Fannie Mae and Freddie Mac and trained on structured mortgage datasets, including Uniform Loan Application Dataset (URLA) files and bank transaction records."

"Forty-two percent on compliance questions should worry every lender relying on off-the-shelf AI right now, Yu said. When I was building technology at Better.com, I watched general-purpose tools fail on mortgage data over and over. They'd miss a payroll mismatch or hallucinate a deposit source, and a human had to catch it every time."

Tidalwave's SOLO AI system demonstrates substantial performance advantages over generic large language models in mortgage lending applications. SOLO achieved 95% accuracy on yes-or-no compliance checks compared to 42% for baseline models, identifying issues like payroll mismatches and suspicious transactions. While Claude 4.5 outperformed SOLO on account verification (86% vs 67%), SOLO excelled in transaction identification (83% vs 80%). The performance differences stem from SOLO's integration with underwriting systems from Fannie Mae and Freddie Mac, training on structured mortgage datasets including URLA files and bank records, and removal of personally identifiable information before processing. Generic models treat loan files as text and risk violating customer privacy by processing PII data. Loan officers increasingly adopt AI tools to manage origination timelines exceeding 40 days, as lenders face pressure to automate processes due to thin margins.

Read at www.housingwire.com