Waterstone Mortgage Taps Friday Harbor Ai To Boost Underwriting Capacity

Friday Harbor, an AI-powered mortgage origination platform, said Wednesday that national retail lender Waterstone Mortgage has implemented its pre-underwriting technology to improve loan file quality and expand underwriting capacity across the company’s nationwide retail channel.

Waterstone Mortgage, which said it has more than 200 loan officers operating in 48 states, deployed Friday Harbor’s platform to surface documentation and income issues earlier in the loan process. It also seeks to deliver more complete files to a relatively lean underwriting staff, according to the company announcement.

Since rollout, Waterstone has reported stronger preapproval confidence, faster file resolution and fewer conditions at the underwriting desk, according to the announcement. That combination is increasingly important for lenders trying to grow production without adding fulfillment headcount in a market still marked by margin pressures and uneven volumes.

“Cleaner files mean our underwriting team can focus on high-value risk decisions instead of clearing avoidable conditions,” Jake Rowoldt, vice president of information services at Waterstone Mortgage, said in a statement. “Friday Harbor is the first AI solution we’ve tried that actually delivers on what it promises. It’s helping us scale production while maintaining the standards that define our business.”

Friday Harbor’s software analyzes loan files during origination to flag missing documents, surface income calculation issues and identify potential underwriting conditions. The tool compares files against investor guidelines and lender overlays, giving loan officers visibility into problems that might otherwise appear days or weeks later in the process.

The platform is integrated with Waterstone’s Encompass loan origination system from ICE Mortgage Technology, allowing loan officers to access the AI assistant within existing workflows. Friday Harbor said Waterstone piloted the technology in late 2024 and has since expanded it across sales and operations teams. Loan officers report that the system’s income and asset calculations closely align with their own and with underwriting’s final decisions.

“Waterstone approached this implementation with a clear understanding of what they needed: a way to improve file quality without disrupting their workflows or undermining the expertise of their underwriters,” Friday Harbor CEO Theo Ellis said. “Their success demonstrates what becomes possible when lenders use AI to support judgment rather than replace it.”

The move underscores how lenders are testing AI in narrow, operationally focused use cases such as income calculation, conditions management and documentation checks rather than end-to-end underwriting. For mortgage executives, the value proposition centers on reducing touches per file and cutting manufacturing defects while staying within current regulatory expectations around fair lending, data use and model governance.

As secondary market investors and regulators continue to scrutinize underwriting quality and repurchase risk, tools that can help lenders reduce underwriting conditions and improve preapproval accuracy are drawing attention. For retail lenders with multistate footprints and distributed sales teams, AI-enabled pre-underwriting can be one way to standardize file quality without significantly expanding underwriting staff.

More details on the Waterstone implementation are available in a case study published by Friday Harbor.