A credit union’s first real line of defense
They had core banking and basic rules. What they didn’t have was real-time fraud detection and loan offers that felt personal.
They had core banking and basic rules. What they didn’t have was real-time fraud detection and loan offers that felt personal.
Fraud was caught after the fact — if at all. Loan products were one-size-fits-all; members with strong history still saw generic rates. The board wanted to compete with digital-first players but didn’t want to rip out the core or hire a 50-person data team.
We built two layers that plug into their existing core. First: a real-time transaction monitor that scores each payment and login using behavior and device signals. High-risk events get flagged for review or step-up auth before money moves. Second: a recommendation engine that uses anonymized member behavior and eligibility to surface the right loan and rate at the right time in their app and in-branch tools.
Fraud losses dropped sharply in the first year; ops spends less time on false positives. Loan uptake from recommended offers increased, and members started describing the experience as “finally feeling seen.” The credit union kept their core, their culture, and gained a story they could take to the next annual meeting.