Teachers Federal Credit Union has reduced its loan decision time by 75% with deployment of Corridor Platforms’ gen AI-powered underwriting tool. 

The $9.9 billion credit union started working with the automation service provider in the first half of 2024 to enhance member experience, improve decisioning speed and agility, drive higher member growth digitally and unlock the full potential of member data, Brad Calhoun, chief executive at Teachers Federal Credit Union, told Bank Automation News.

“Teachers booked approximately 30,000 loans for fiscal 2024 [which ended Dec. 31],” Calhoun said. “A majority of these loans are related to automobile loans, credit cards, personal loans and lines of credit which are now being decisioned by models and rules developed, tested and maintained in Corridor’s Smart Decisioning Platform [SDP].” 

The SDP was integrated into the credit union’s production workflow system via APIs, which included the loan origination system for real-time decisioning, he said. The SDP was also connected to Teachers’ in-house data lake to build decisions analytics based on Teachers FCU risk appetite. 

“Teachers FCU did not need to do any tech upgrades to deploy Corridor’s advanced decisioning tools,” Calhoun said. 

AI decisioning 

Automated lending decisioning automatically approves or denies loans without needing a human underwriter to review the decisions, while staying within the FI’s risk guardrails, Manish Gupta, CEO at Corridor Platforms, told BAN. 

The AI model looks at three main criteria when reviewing an application: 

• FICO scores; 

• Post risk assessment, or length of employment, affordability metrics like existing debt burden; and 

• Loan-specific metrics such as loan-to-automobile value. 

“This decision needs to be automated in as few as 100 milliseconds in the digital marketplace to be competitive against Tier 1 and fintech lenders,” Gupta said, adding that the AI model built in-house can perform the underwriting quickly and accurately. 

The Corridor Platforms AI decisioning model can be customized according to FI’s risk appetite, Gupta said. The model “generates a locked artifact after approval on the platform” so that reviewers can safely look at why a loan was approved. 

Corridor’s clients include two of the top five consumer banks and a digital lending fintech, Gupta said, declining to disclose the names of the FIs. 

FIs using gen AI-driven underwriting include: 

• $3 trillion HSBC; 

• $1.1 trillion Lloyds; 

• $607 million OneUnited Bank; and 

• $659 billion U.S. Bank. 

Corridor is expanding platform features with GenGuardX, a governance framework for deploying gen AI applications like chatbots and virtual agents in production environments, Gupta said.