What does AI in mortgage really look like in action? Rebecca Seward joins Diego Sanchez to explain how AI isn’t here to replace underwriters or loan officers — it’s here to amplify them. She shares insights on how AI can streamline processes, reduce repetitive tasks and help mortgage professionals focus on higher-value work. 

Rebecca also explores how AI can help create long-term industry stability by disrupting the traditional boom-and-bust workforce cycle and supporting a more sustainable staffing model. This conversation breaks down practical AI strategies, showing how the right tools can provide better insights and help teams work smarter — not harder. 

AI as an augmentation tool

“AI is essentially an augmentation tool that enables lenders to make faster, better decisions,” Seward explained. “We do not see ourselves as a replacement for the human decision-makers. We believe that we have a part to play to enable them to be more efficient.”

Ocrolus uses a combination of artificial intelligence, machine learning, and optical character recognition (OCR) to provide lenders with decision-ready information. However, Seward emphasized that humans remain central to the process. “The AI is only as good as the information that you give it,” she said. “We use human validators to constantly improve our systems, making our analytics smarter and better over time.”

Scaling efficiently

“Every time we see that refi boom or the surge in volume, lenders respond by hiring a ton of human collateral,” Seward noted. “If you embed technology like Ocrolus, it enables you to scale so much more efficiently. You train your team to be 3x to 4x more productive, process more loans, and make more accurate decisions.”

“We extract and classify 1,700 different document types, standardizing them across all lender workflows,” Seward said. “For example, when we receive a bank statement, we automatically label and categorize it according to the lender’s unique taxonomy, removing the manual work for processors.”

Ocrolus provides advanced analytics, including income calculations across various borrower types. “We handle wage earners, self-employed borrowers, rental income, and even variable income sources like unemployment and retirement,” she explained. “We’re also piloting a new tool called Inspect, which compares data from the 1003 application with the actual documents to flag inconsistencies, like undisclosed REO or missing documents.”

Reducing the boom-and-bust hiring cycle

“By enabling teams to work more efficiently, you don’t have to go through massive hiring and firing cycles,” Seward said. “You can handle surges in volume without drastically expanding your workforce, and when the market slows, your team is still operating at a high level without unnecessary overhead.”

A flexible AI strategy for lenders

“We involve our human teams in decision-making and continuously improve our models based on user feedback,” Seward explained. “We train underwriters on AI-powered tools, ensuring they know how to use them to their advantage rather than fearing replacement.”

Seward also believes AI will be critical for attracting the next generation of mortgage professionals. “New talent—especially Gen Z and Millennials—expects technology to be part of their workflow. They’ve grown up with tools like ChatGPT and will be looking for companies that embrace AI rather than resist it.”

The future of AI in mortgage

“We aim to get loan processors and underwriters 90% of the way to a decision, but the final step is always in human hands,” Seward said. “We’re not here to replace people—we’re here to make them better and help them do their jobs more effectively.”