Vesta Raises $30M Series B to Automate Mortgage Underwriting Tasks With AI Agents

Mortgage technology startup Vesta has raised a $30 million Series B funding round to expand artificial intelligence agents that process loan origination workflows, as reported by The Next Web (https://thenextweb.com/news/vesta-30m-mortgage-ai-agents-eu). Conversion Capital led the investment, with participation from Citi Ventures, Andreessen Horowitz, and three mortgage lenders currently using the platform.

The round brings Vesta's total funding to $85 million. Chief executive Mike Yu told TechCrunch that annual revenue increased twelvefold over the past year amid growing customer demand, while noting the company still holds less than 5% market share. Lenders using the platform originate more than $100 billion in mortgages annually, according to HousingWire, with automated workflows and AI agents now managing roughly 40% of tasks processed by the system.

The platform's agents handle document reviews, data validation, condition clearing, closing package reviews, and some underwriting decisions. The software aims to compress mortgage processing timelines and costs in the United States, where closing an origination takes an average of 40 days and costs roughly $11,000—expenses driven primarily by labor, according to data from the Mortgage Bankers Association.

Yu credited Anthropic's Claude Sonnet 4.5 model with enabling agents to adhere to configured instructions across complex, multi-step origination processes. Addressing scrutiny over automated bias—highlighted by a 2021 investigation by The Markup that found US lenders were 80% more likely to deny Black applicants than comparable white applicants after controlling for 17 financial factors—Yu stated that lenders maintain direct accountability for underwriting decisions, noting that Vesta logs all system actions and reasoning for audit trails.

The expansion of automated lending tools highlights sharp divergence in international regulatory requirements. Under the European Union's AI Act, systems that evaluate the creditworthiness of natural persons are classified as high-risk, requiring provider conformity assessments, mandatory human oversight, six months of activity logging, and fundamental rights impact assessments. While Regulation (EU) 2026/1744 deferred those obligations to Dec. 2, 2027, Article 22 of the General Data Protection Regulation already restricts decisions based solely on automated processing that carry significant legal effects, including mortgage rejections.