Fiserv achieves 98% end-to-end automation for merchant category code validation with agentic AI

In payment processing, merchant category codes (MCCs) are foundational — they govern interchange rates, fraud risk scoring, rewards eligibility, and regulatory compliance for every transaction processed. For Fiserv, a global fintech serving banking, merchant acquiring, and commerce clients at scale, validating and updating MCCs had become a significant operational bottleneck. Customer service representatives handling merchant requests spent 10–15 minutes per call manually extracting data from multiple disconnected systems, managing separate application logins, and transcribing information by hand. The process was error-prone, extended resolution times across thousands of daily interactions, and complicated onboarding for new hires who had to master a fragmented stack of legacy applications before contributing productively.

Fiserv deployed UiPath's agentic automation platform to rebuild MCC validation end-to-end. UiPath Robots pull merchant records directly from Fiserv's internal database, cross-reference business details through a Bing API lookup, and apply UiPath GenAI Activities to select the correct MCC via generative AI prompts — processing standard cases without human input. A rules engine now resolves 98% of requests automatically; exceptions are escalated to UiPath Action Center, where a CSR can review and approve without leaving a single interface. On the front end, UiPath Apps consolidated fragmented system logins into one unified workspace and enabled in-app note-taking that triggers downstream automation bots automatically. UiPath Autopilot handles workflow troubleshooting and is being piloted for contract and SLA use cases as Fiserv scales the platform enterprise-wide.

The MCC automation program saves Fiserv 12,000 hours annually, eliminating the manual effort that previously fragmented every merchant call across disconnected systems. The workflow now achieves a 98% end-to-end automation rate, with only a small fraction of edge cases routed to human review.

• 12,000 hours/year reclaimed from manual MCC validation
• 98% of MCC cases resolved without human intervention
• CSR call handling time cut from 10–15 minutes to near-zero for standard requests
• New hire onboarding simplified — staff now train on one UiPath Apps interface rather than navigating multiple legacy systems

These results provided the measurable ROI Fiserv needed to justify accelerating agentic AI adoption across additional departments and workflows.

• Anchor enterprise AI expansion on a single high-visibility use case with clear, quantifiable ROI — success there builds the internal trust needed to scale.
• A layered architecture combining RPA robots, generative AI prompts, and structured human-in-the-loop review is more resilient than full automation alone; design exception routing before go-live.
• Simplifying the employee-facing interface matters as much as back-end automation — reducing login friction and consolidating workflows directly accelerates adoption and cuts onboarding time.
• In payment operations, MCC accuracy has compounding downstream effects on interchange fees, fraud models, and compliance — operational automation here delivers value well beyond headcount savings.