Valon Raises $150M to Scale AI Mortgage Servicing

Valon spent six years operating a mortgage servicer so its software could learn the part of regulated finance that never fits inside a product demo: payment histories, escrow changes, investor reporting, compliance decisions and the exceptions that accumulate across a 30-year loan. The company has now raised a $150M Series D at a $2.3B valuation to move that operating knowledge into more of the mortgage industry.

Ribbit Capital joined as a new investor, while Andreessen Horowitz participated again. Valon says the financing will accelerate ValonOS product development, expand engineering, product, deployment and go-to-market teams, and support migrations at some of the country's largest mortgage servicers. The broader wager is that applied AI in regulated finance will be limited less by model intelligence than by whether the underlying system can preserve context, execute deterministic actions and explain what happened after the fact.

What Happened

Valon announced the Series D on October 5, 2026. The company said the $2.3B valuation doubled its prior mark, and the financing follows a $100M Series C led by WestCap in October 2024. Valon's earlier financing history also includes a $50M Series A led by Andreessen Horowitz in 2021, a $43.9M investment later that year and an undisclosed long-term minority investment from Rithm Capital announced in February 2026.

The latest announcement names Ribbit Capital as a new investor and Andreessen Horowitz as a returning participant, but it does not identify a lead investor or disclose the complete syndicate. Valon says the new capital will go toward product development and hiring in New York, San Francisco and remote roles as it deploys ValonOS and its AI agents across more large servicing operations.

The Operating Business Became the Product Lab

Mortgage servicing begins after a loan closes. It covers payment accounting, escrowed taxes and insurance, investor remittances, borrower assistance, defaults, foreclosures and the regulatory record surrounding each action. Those responsibilities forced Valon to build more than a cleaner interface because a system of record has to carry both the numbers and the reasoning behind each exception.

Valon built and ran a licensed servicer on ValonOS for six years. In August 2026, Carrington Mortgage Services completed its acquisition of Valon Mortgage, adding approximately 810,000 loans to Carrington's servicing operation and adopting ValonOS as its core platform. That transaction let Valon Technologies focus on the software while keeping its product tied to operating conditions that had already tested payments, escrow, compliance and customer support at scale.

Current leadership includes CEO and co-founder Andrew Wang, President, COO and co-founder Linda Du, and co-founder Jonathan Hsu. The company's official leadership page also lists Brian McGrath as Executive Chair, Hongxia Zhong as Head of Engineering, Jake Mintz as CPO, Marisa Rackson as Head of People and Ariel Brito as Chief of Staff. Valon does not currently identify a CTO on that page.

Why ValonOS Is More Than an AI Layer

ValonOS combines loan data, accounting ledgers, investor reporting, servicing workflows, task management, compliance logic and money movement in one system of record. The company says the platform preserves the exceptions, precedents and decision traces behind servicing activity, giving human operators and AI agents the same source of truth and guardrails.

That architecture matters because mortgage servicing is full of actions that must be both correct and explainable. An agent responding to a payment increase may need the prior escrow analysis, current tax information and the history of an approved exception before it can produce a useful answer. Valon's native agent, Ditto, is designed to work inside that context across homeowner communications, escrow analysis, complaint research and payment reallocations.

The distinction is practical rather than theatrical. A model can generate language in seconds, but a servicer still needs controlled access to financial records, deterministic execution and an audit trail that can survive internal review or regulatory examination. Valon's financing supports the expensive work of moving those capabilities from a company-operated proof environment into customer systems that already hold decades of obligations.

Customer Commitments Turn Product Proof Into Migration Risk

Valon reports more than $200B in mortgages already running on ValonOS and more than $200M in contracted annual recurring revenue signed within the first six months of opening the platform to outside servicers. It also says institutions responsible for 1 in 6 outstanding U.S. mortgages are under contract to use ValonOS. Those figures are company-reported and have not been presented as independently audited results.

Two of the ten largest U.S. servicers, Carrington and ServiceMac, are already live on the platform, according to the funding announcement. Rithm Capital's February partnership release says Newrez plans to begin moving more than 4M homeowners onto ValonOS in 2027. Together, those commitments make implementation quality the next visible measure of the company.

Replacing a core servicing platform is not a normal software rollout. Data has to reconcile, decision history has to survive, investor reports have to remain accurate and customer service cannot disappear while a migration team rebuilds the machinery underneath it. The Series D gives Valon more capacity to absorb that work, but the credibility of the platform will be earned inside those handoffs.

What the Series D Signals for Regulated AI

Valon's trajectory reflects a larger enterprise-AI shift. The most valuable agent deployments in regulated industries may depend on rebuilding or replacing the system of record rather than placing a conversational layer over software that cannot expose its own logic. Andreessen Horowitz has framed this challenge as turning regulation into code and giving agents structured workflows, auditable ledgers and programmable actions.

Valon eventually plans to apply its architecture beyond mortgage servicing to commercial, personal, auto and student lending. That expansion remains a stated ambition, not an achieved outcome, and each category carries its own rules, data models and operating exceptions. The mortgage market is already large enough to test whether the same platform can move from two major live servicers to a meaningful share of the industry without losing the correctness that made the original proof useful.

The $150M round does not remove that implementation burden. It funds Valon to meet it at a larger scale, with more customers and less room for any exception to fall between the old ledger and the new one.