Delivering AI-Driven Loan Decisioning to Credit Unions

By Roy Urrico

Artificial intelligence (AI)-driven decision automation is transforming financial institutions (FIs) that must deliver highly personalized interactions by accessing data quickly. Corridor Platforms with its advanced decision and analytics workflow automation platform, has helped Teachers Federal Credit Union boost its loan approvals and improve real-time decision automation with its RiskDecisioning.ai (RD.AI) solution. 

Manish Gupta, CEO and co-founder of Corridor Platforms.

Manish Gupta, CEO and co-founder of Haworth, N.J.-based Corridor Platforms, and Brad Calhoun, CEO of the Hauppauge, N.Y.-based Teachers, discussed with Finopotamus AI-driven decision automation’s impact at FIs.

“Mid-market financial institutions today face a critical challenge: Consumers now expect instant, seamless, and highly personalized experiences. For banks and credit unions, meeting these expectations requires leveraging vast amounts of data about current and prospective customers in real time. Yet, the very data needed to meet these demands is equally accessible to competitors,” Gupta said.

This creates a serious disadvantage for financial institutions constrained by outdated systems, fragmented data, and limited decisioning capabilities, Gupta explained. “These limitations leave [mid-market FIs] struggling to compete effectively with tier-one banks and fintech disruptors that are already leading the charge in personalized, data-driven decisioning.”

Brad Calhoun, CEO of Teachers Federal Credit Union.

Gupta described how RD.AI enables turnkey digital lending programs for mid-size banks and credit unions. “We enable them to leverage advanced analytics, AI, and the cloud to deliver instant credit decisions necessary to support digital lending initiatives such as point-of-sale financing, digital loans, mortgage approvals, and real-time credit limit changes.”

“By integrating Corridor Platforms’ real-time credit decisioning capabilities, Teachers FCU aimed to elevate the member experience, optimize risk management and governance through automation, and catalyze growth at an unparalleled rate,” acknowledged Calhoun.

Enabling Turnkey Lending

The inability to deliver best-in-class decisioning results in sub-par customer relationships, increased portfolio risk, and the potential loss of valuable customers to more agile competitors, offered Gupta. “Mid-market FIs must move beyond rented point scores and instead build robust in-house capabilities that allow them to make contextually relevant, instant decisions while maintaining compliance with stringent lending laws.”

Gupta explained that Corridor’s RD.AI development team formed after spending decades in financial services managing risk across credit cycles, designing innovative lending products, and overseeing large profit and loss (P&L). “We saw firsthand how difficult it is for regulated entities to enhance decisioning and analytics in the face of rapid innovation,” he continued. “The existing decision management workflow involves many manual handoffs between data teams, modeling teams, strategy implementers, and governance groups, leading to long lead times in developing, testing, and implementing any change in decisioning logic.”

The more sophisticated the data, models, and strategies, the longer it takes most FIs to go through governance and testing procedures for implementation in production, Gupta said. “In the digital marketplace, the pace of change in data, AI, and GenAI (generative AI) is so fast that most financial institutions struggle to keep up with agile competition. We aimed to develop end-to-end digital decisioning capabilities that allowed regulated entities to innovate using cutting-edge AI and big data technologies on a highly automated and governed platform,” he added.

Another driving factor behind building RD.AI, recalled Gupta, “was the slow pace at which FIs iterated on analytics and the inefficiencies they faced in deploying them to production. Our platform enables this by providing end-to-end connectivity, automating and compressing the change management lifecycle from weeks to hours.”

Making Digital Lending Turnkey

At the heart of RD.AI is the ability to seamlessly connect data, modeling, and decision strategy workbenches on one “rail,” enabling FIs to adopt the latest modeling advancements without operational disruption.

Gupta further explained benefits of the solution. “We enable FIs to build and own their core in-house capability instead of just renting it for long-term value, and we span across the entire customer lifecycle, be it origination, marketing, cross-selling, or loss management. This is critical. We want to optimize the full member lifecycle, not just the initial selection at the acquisition point.”

A key strength of RD.AI’s offering is its modular design, allowing FIs to tailor its use to their specific needs and adopt a "think big, act small, and move fast" mindset, said Gupta. “For instance, an FI can start small by utilizing the platform to refine credit strategies, rigorously test them, automate compliance approval workflows, and maintain ongoing monitoring, while keeping a big vision for transformation.”

Teachers’ Transformation Journey

The $9.86-billion Teachers Federal Credit Union embarked on the transformation journey toward member decisioning with RD.AI in the first half of 2024. Its goal was to enhance the experience, improve decisioning speed and agility, drive higher growth digitally and unlock the full data potential for its approximately 468,000 members. “Within 4-5 months post engagement, the decision analytics for models and strategies were configured and onboarded on Corridor and Teachers went live in the last quarter of 2024,” Calhoun told Finopotamus.

“A top priority was to ensure that Teachers could extend financial products and services to more members,” Calhoun said. “Teachers sought to harness Corridor’s deep experience with major financial institutions while maintaining full control over their data and core operations.”

Calhoun continued, “Corridor’s proficiency in automating decisioning processes aligns seamlessly with Teachers’ commitment to surpassing member expectations through innovative digital solutions and by leveraging their rich internal data resources. This strategic partnership not only enhances operational efficiency but also drives cost-effectiveness through the automation of analytics, reporting, and real-time decision-making.”

Corridor Help Teachers With AI-Driven Decisioning

Calhoun confirmed, “Teachers Federal Credit Union has elevated the sophistication of its decision-making processes across the entire member lifecycle with the support of Corridor’s RD.AI. This transformation, he stated, “is driven by a multi-faceted approach: leveraging internal advancements in data lake management, implementing [RD.AI], upskilling staff to develop more predictive risk models using the platform’s robust data and tools, and prioritizing investments in compliance and governance to ensure responsible deployment of underwriting policies and models.”

The Corridor platform integrated into the Teachers FCU production workflow system, which included the loan origination system (LOS) for real-time decisioning. Custom risk models were built for automobile, credit card, and personal loan portfolios, and current credit strategies streamlined to support optimal decision automation. “Teachers views these upgrades as essential to delivering exceptional service to members and prospects while staying competitive in the rapidly changing digital banking landscape,” said Calhoun.

Since launching, Teachers FCU has boosted its loan approval rates by up to 50%, improved real-time decision automation by over 300%, and created a seamless, enhanced member experience, said Calhoun. Teachers has also seen significant reduction in time from application to decision with Corridor’s RD.AI, according to Calhoun. “Processing time reduced from hours to minutes across the credit portfolios, with a 75% reduction on average decisioning time for new applications,” said Calhoun. “By streamlining the loan approvals, Teachers has significantly improved processing speed and reduced operational burden, all while maximizing the existing resources. The impact has been staggering—had Teachers implemented this model earlier, they could have reduced charge-offs by nearly 30%.”

Corridor Differentiators

Gupta said Corridor Platforms’ RiskDecisioning.ai presents two distinct and connected modules:

1. Data and Model Management. Connecting the data, modeling, and business strategy workbenches (teams and their tools), eliminates manual handoffs, reduces operational inefficiencies, creates transparency across the decision lifecycle, and provides auditability that allows for enhanced controls.

2. Strategy Automation. Strategies are evaluated, sent for approvals, programmed by technology, and tested before put into production. Corridor created end-to-end connectivity that enables automation and compresses the change management lifecycle into hours rather than weeks or months.

Other key differentiators include:

• End-to-end digital decisioning lifecycle solution. That ‘strings the pearls’ for decision objects of data, features, models, and strategies/rules in a controlled, interconnected, shared digital workbench.

• System of record for analytics. The platform becomes the ‘single source of truth’ for all analytics at a regulated firm, including analytics use cases, which were approved and locked on the platform for post-production scrutiny.

• Systematic governance and compliance. It has become imperative for FIs to maintain full control over their decisioning processes to ensure regulatory confidence.

• One click-to-production. AI-powered decision artifacts move from the analytical environment to production post compliance and approval in 'one-click' with no recoding.

According to Gupta, the RD.AI platform connects to all major loan origination and core banking systems in the marketplace. The decisioning platform can reside on-premise at the FI, on the FI’s private cloud, or on Corridor’s secure SOC2-compliant cloud.