28-07-2025
Exclusive: Loan servicing startup Salient raises $60M Series A
Salient, a San Francisco-based platform for lenders to automate the post-loan origination process, raised $60 million in Series A funding, it tells Axios Pro exclusively.
Why it matters: Loan servicing remains highly manual, but banks are increasingly doubling down on tech in the area.
Andreessen Horowitz led the round, joined by Matrix Partners, Michael Ovitz and Y Combinator.
By the numbers: The deal values Salient at $350 million, Axios has learned.
The company's annualized run rate was north of $14 million as of June 2025, some 18 months after its launch.
Context: This comes as U.S. household debt hit $18.2 trillion in the first quarter, with 4.3% of that delinquent — the highest level in five years, according to the New York Fed.
How it works: Salient uses generative AI to automate collections, customer service, and compliance monitoring — acting as a dashboard for lenders to track all their loans.
Currently, "[lenders] have huge outsource firms that run compliance functions or call centers," says CEO Ari Malik. "What we're trying to offer is a more transparent way of seeing what's happening. Because if you outsource this whole process, oftentimes, lenders have no idea what's going on."
Salient uses voice recognition to monitor customer service calls and flag violations of complex state or federal lending rules. For example, active service members are entitled to interest rates of 6% or below, and customers who wish to no longer be called must be marked. Failures can lead to steep fines.
In addition to customer service, Salient's AI agent, with additional information from the customer, can complete insurance claims and get necessary paperwork from the lender — cutting down on time and complexity.
Salient counts Westlake Financial and AutoNation among its customers.
Zoom out: Malik is betting this automation can improve lender and customer experiences.
"[The AI agent] needs to be predictive as to what the customer wants," he says. "Success is: 'Can you predict what they actually want to interact with you about, as opposed to starting everything from scratch?'"
