AI Agent Swarms Now Process $100B in Mortgages a Year

Vesta is a startup that builds AI agent swarms for mortgage lenders. It just raised $30 million in a round led by the investment fund Conversion Capital. That brings the company's total funding to $85 million. Vesta's agents carry a mortgage loan from document intake through the underwriting decision. The system now processes more than $100 billion in loans per year. CEO Mike Yu said revenue grew 12x year over year. He credited most of that jump to Claude Sonnet 4.5, the Anthropic model that powers Vesta's agents. Yu said it follows a lender's instructions more precisely on long tasks than earlier models did.

In the US, originating a mortgage takes about 40 days and costs roughly $11,000 per loan. Most of that is human labor spent waiting for someone to review the file. Holding onto an instruction without losing the thread across a stretch that long is what separates a working agent system from a demo. It fits a broader pattern, too. Competing models, including Chinese ones, tend to drift on long tasks. Anthropic's models keep following instructions through to the end.