Vesta, an AI-native mortgage technology startup, has raised $30 million in a funding round led by Conversion Capital as it looks to expand its AI-powered mortgage origination platform and introduce new products for lenders.

The round also included investments from Citi Ventures, Andreessen Horowitz (a16z), and three of Vesta's customers, including Pennymac and New American Funding. The latest investment brings Vesta's total funding to $85 million.

Founded in 2020 by Mike Yu and Devon Yang, Vesta develops software designed to automate mortgage origination, the process lenders follow to evaluate applications, review documentation, assess borrower eligibility, and move loans toward closing.

The company is building an AI-native alternative to traditional mortgage systems, using teams of AI agents to handle complex, multi-stage workflows that have historically required substantial human involvement.

Vesta targets the cost and delays of mortgage lending

Mortgage origination in the United States remains a time-consuming and expensive process. According to Vesta CEO Mike Yu, closing a mortgage typically takes around 40 days and costs approximately $11,000 per loan.

Much of that expense comes from human labour, while delays often occur when loan applications are waiting for someone to review documents or complete the next step.

Vesta aims to reduce these bottlenecks by allowing lenders to assign specific tasks to AI agents. Rather than requiring employees to manually manage every stage of a loan, the platform can automate selected activities and move applications through workflows more efficiently.

Lenders retain control over how much responsibility they give the software. A company can initially require employees to approve an agent's work, then gradually allow the system to handle a larger share of loans independently as confidence grows.

According to Yu, some customers are already using Vesta's AI agents to make mortgage underwriting decisions. However, lenders remain responsible for those decisions, regardless of the software they use.

Vesta records the actions and reasoning behind its agents' decisions to support compliance and auditing. This is particularly relevant in mortgage lending, where decisions must be explainable and organisations need to maintain oversight of their processes.

Why AI agents are becoming more practical

Vesta's approach reflects a broader shift in how AI is being applied to enterprise software.

Earlier generations of AI models were not reliable enough for the extended, multi-stage tasks involved in mortgage origination. The company initially concentrated on developing the data architecture needed to support automation using increasingly capable AI systems.

Yu identified Anthropic's Claude Sonnet 4.5 as a major breakthrough for Vesta, saying it followed user-configured instructions over longer periods more effectively than previous models.

That improvement matters because mortgage workflows involve multiple connected tasks, changing information, strict rules, and decisions that may depend on earlier steps. An AI agent must retain the relevant context and follow the lender's instructions throughout the process rather than simply respond to an individual prompt.

Vesta's platform is designed to let lenders configure agents around their own processes and gradually expand their responsibilities. The goal is to automate more of the work without removing the human oversight required for regulated lending.

Revenue grows 12x as demand increases

The latest funding comes as Vesta reports a sharp increase in demand for its software.

Yu told TechCrunch that the company's revenue has grown 12 times year over year. Despite that growth, he said Vesta still serves less than 5% of its addressable market, leaving significant room for expansion.

The company plans to use the new capital to increase hiring, deepen its reach across the mortgage industry, and develop additional products.

One planned product is a personal assistant for mortgage issuers. It is intended to help lenders perform tasks and track workflows, extending the platform beyond individual loan-processing activities.

Vesta's strategy is to win more business from mortgage lenders before expanding into other areas based on customer demand.

Competition with established mortgage technology providers

Vesta operates in a market that includes both established mortgage software providers and newer AI-focused startups.

Its competitors include ICE Mortgage Technology, a major provider of mortgage technology, and Xpanse, which is also working to automate mortgage lending.

Vesta argues that AI-native software gives it an advantage over legacy platforms that were not originally designed around autonomous agents. Adding agents to older systems can be difficult when their underlying architecture was built for conventional software workflows.

The company is betting that lenders will increasingly prefer systems designed from the ground up to support AI-driven operations.

Still, broader adoption will depend on whether these systems can deliver measurable savings while maintaining accuracy, security, compliance, and appropriate human oversight. Mortgage underwriting carries significant financial consequences, making reliability as important as speed.

What the $30M funding means for Vesta

With $30 million in fresh capital and $85 million raised to date, Vesta is preparing for its next phase of growth. Its immediate priorities include expanding its customer base, hiring more employees, and developing new AI products for mortgage issuers.

The opportunity is substantial: reducing delays and manual work across a process that can take around 40 days and cost roughly $11,000 per loan could create meaningful operational benefits for lenders.

The challenge will be proving that AI agents can manage increasingly complex workflows consistently and responsibly at scale.

For Vesta, the next stage is about turning strong revenue growth and early customer adoption into a larger position in the mortgage technology market.