Vesta Secures $30 Million to Deploy AI Agent Swarms for Mortgage Lenders

Mortgage technology startup Vesta announced on Thursday that it has closed a $30 million funding round led by Conversion Capital. The capital will support the expansion of the company’s AI-native platform, which utilizes swarms of autonomous agents to automate various stages of the loan origination process.

The investment round features participation from existing clients Pennymac and New American Funding, alongside strategic investors Citi Ventures and Andreessen Horowitz. This brings Vesta’s total funding to $85 million since its founding in 2020 by Mike Yu and Devon Yang.

Yu, who serves as CEO, described recent demand for the company’s services as having “exploded.” He reported that revenue has increased twelve-fold year over year and that Vesta has facilitated over $100 billion in annual loan originations for its lending partners. Despite this traction, Yu noted that the company currently holds less than 5% market share, stating that the timing is ideal to hire staff and develop new product lines.

Among the planned innovations is a personal assistant tool designed to help mortgage issuers manage workflows. The U.S. mortgage closing process typically takes around 40 days and costs approximately $11,000 per loan, with human labor representing the majority of expenses and significant delays caused by manual review bottlenecks. Vesta’s model allows lenders to deploy agents to accelerate these tasks.

“Many of our customers start an AI agent with a person approving its work, then let it handle a share of loans on its own, then expand,” Yu explained. He added that some lenders are already using the agents to assist with underwriting decisions. However, he emphasized that the lending companies retain full responsibility for those decisions. All actions and reasoning processes are logged to ensure compliance and allow for auditing.

Yu attributed the ability to build such autonomous systems to recent advancements in large language models, specifically citing the capabilities of Claude Sonnet 4.5. He noted that earlier models lacked the ability to adhere to complex, user-configured instructions over the long time horizons required for mortgage processing. Previously, Vesta focused primarily on establishing robust data architecture to support future automation.

The company operates in a competitive landscape facing both legacy providers like ICE Mortgage Technology and other AI-focused entrants such as Xpanse. Yu argued that Vesta holds a distinct advantage over incumbents because traditional systems were not engineered for AI agents, making integration difficult. Looking ahead, Yu stated the priority is to capture a larger share of the mortgage industry, adding, “Then, we’ll go wherever our customers take us.”