Digital homeownership company Better Home & Finance Holding Company (Better.com), has debuted Betsy, a voice-based AI loan assistant for the U.S. mortgage industry.
Built through Tinman, the company’s proprietary loan origination platform, Betsy leverages AI and large language models (LLMs) to accelerate a customer’s entire mortgage journey from pre-approval start to closed loan. The platform is programmed to communicate with prospective and existing Better.com customers to answer mortgage application inquiries and to collect and verify outstanding application data, all while interfacing with Tinman in real-time.
Tinman is a single, unified system where all relevant facts of a loan application are stored within one ] data structure. This contrasts with traditional mortgage industry software, where information is spread across multiple systems and datasets such as Point of Sale, CRM, Loan Origination System, Document Management System and Pricing Engine.
“Fulfilling a mortgage loan is labor-intensive and rules-driven, and Betsy serves as a highly intelligent front end enabling customers to automate the process with the full power of our proprietary loan engine, Tinman,” said Vishal Garg, CEO and founder of Better.com. “We believe that Betsy will catalyze our growth over the coming years by making our existing loan officers, processors, coordinators and closers far more efficient.”
Tinman’s centralized and context-rich data environment is organized in a hierarchical, tree-like structure that is easily understood by LLMs like Betsy. This allows Betsy to access and interpret information with full context, enabling her to understand the nuances of each customer’s loan application. As a result, Betsy can accurately answer questions in great detail and assist with outstanding tasks efficiently. The results enable faster service times, enhanced self-service capabilities, improved customer engagement, and greater sales efficiency.
With access to a repository of customer-provided application data, Betsy enables loan officers to focus their time on discussing interest rate details and other licensed activity with a seamless transition of information.
“The average cost to sell and process a mortgage in the United States is nearly $9,000 per loan,” said Kevin Ryan, CFO of Better.com. “Utilizing Tinman’s capabilities, we have been able to automate time and labor-intensive components of the mortgage process and reduce that cost by over 35% of the industry average. As we continue to supplement Betsy’s capabilities, we believe this technology will significantly drive down our costs further, resulting in lowered rates and superior service for our customers.”
