Mortgage software startup Vesta announced on Thursday that it raised a $30 million funding round led by Conversion Capital. The company builds AI-native software to help lenders originate mortgages.
Mike Yu and Devon Yang co-founded the company in 2020. Yu serves as the CEO. Three of the company customers, including Pennymac and New American Funding, participated in the round alongside Citi Ventures and Andreessen Horowitz.
Yu told TechCrunch that demand for the product exploded in the last year and revenue is up 12x year over year. The company has raised $85 million in total funding to date. Even with this growth, Yu noted that the firm holds under 5% market share.
The company plans to use the new capital to staff up, capture more of the market, and invest in new product lines. These new products include a personal assistant for mortgage issuers to perform tasks and track workflows.
Read nextChina AI Lab Manus Raises Over $500 Million in First Funding After Meta SplitHuman labor drives high mortgage costs
Closing a mortgage in the United States takes around 40 days and costs about $11,000 per loan. Yu stated that human labor drives most of that cost. The timeline also suffers bottlenecks while waiting for people to review loans.
Vesta addresses this by letting humans deploy a swarm of agents to complete tasks faster. Customers choose which tasks to assign to the agents. Lenders often start by having a person approve agent work before letting agents handle a share of loans independently.
Some lenders use Vesta AI agents to make mortgage underwriting decisions. Yu clarified that companies remain responsible for underwriting decisions regardless of the software or AI agents used. The platform records all actions and reasoning to ensure compliance and audit AI decisions.
New AI models enable autonomous agents
The new level of autonomy relies on recent improvements in AI models. Previous generations of AI were not advanced enough to build agents for multi-stage mortgage lending tasks. Vesta previously focused on building data architecture to utilize advanced automation tools.
Yu highlighted Anthropic model Claude Sonnet 4.5 as a major factor for the company. He noted that it adheres better to user-configured instructions over required time horizons compared to older models.
Vesta faces competition from traditional mortgage systems like ICE Mortgage Technology and other AI-native automation companies like Xpanse. Yu argued that legacy incumbents struggle because their systems were not built for AI agents, making it difficult to add AI agents on top.
The company priority for its next phase is earning the business of the rest of the mortgage industry. After that, the company plans to go wherever its customers take it.
