Vesta, an AI-native software company that helps lenders originate mortgages, said Thursday it has raised a $30 million funding round led by Conversion Capital. The startup uses AI agents to automate a large portion of the loan origination process, which its founders say reduces the time and cost of processing mortgages.
Three Vesta customers invested in the round, including Pennymac and New American Funding. Citi Ventures and Andreessen Horowitz also participated.
Mike Yu co-founded Vesta in 2020 with Devon Yang and serves as its CEO. Yu said the timing of the raise reflects demand that has grown sharply over the past year, with revenue up 12 times compared with a year earlier. Vesta has now raised $85 million in total funding.
Yu said that while the company's traction is strong, it still holds less than 5 percent of the market, and that now is the moment to add staff, pursue market share, and invest in new product lines. Among those products is a personal assistant for mortgage issuers that can perform tasks and track workflows. Closing a mortgage in the United States takes about 40 days and costs roughly $11,000 per loan, with most of that cost tied to human labor and much of the delay coming from waiting for a person to review a loan, according to the company.
Vesta's approach is to let lenders deploy a group of agents to complete tasks faster. Customers choose which tasks to assign to the agents. Yu said many customers begin with an agent whose work a person approves, then let it handle a portion of loans independently before expanding further. Some lenders are also using Vesta agents to make mortgage underwriting decisions, he said.
Yu said companies remain responsible for underwriting decisions no matter what software or AI agents they use. He added that all actions and the reasoning behind a decision are recorded for compliance and to allow AI decisions to be audited.
Yu attributed the level of autonomy to improvements in AI models over the past year, saying earlier models were not good enough to support agents for the complex, multi-stage work of mortgage lending. Before that, Vesta focused on building the data architecture needed to automate the mortgage process with advanced tools. Yu said the breakthrough was Anthropic's Claude Sonnet 4.5, which he described as better than previous generations at following user-configured instructions over the time horizons the company needs.
Vesta competes with traditional mortgage systems such as ICE Mortgage Technology and with other AI-native companies working to automate mortgage lending, including Xpanse. Yu said Vesta's advantage over legacy incumbents is that those systems were not built for AI agents, which makes adding agents on top of them difficult. He said the company's priority is winning the business of the rest of the mortgage industry, and that it will follow its customers from there.
More company and startup news from TechManNews.
