AI-Powered Payments Platform — Operational

AI-Powered Payments Platform — Operational Command Center for JP Morgan

About: JP Morgan's Payment Blotter was doing its job managing high-value institutional transactions across multiple teams but nobody could actually see what was happening. Layered permissions, fragmented listing logic, and role-based visibility rules that had accumulated over years meant the system provided data but not clarity. Operations associates, client service teams, and back-office stakeholders were all looking at the same platform and interpreting it differently. In a compliance-driven, time-sensitive environment, that ambiguity isn't a UX problem it's operational risk.

What I did: I was the sole designer on a 9-person cross-functional team, responsible for auditing the entire workflow architecture, untangling role-based logic inconsistencies, and redesigning the blotter into a structured operational workspace while simultaneously laying the AI foundation for future enhancements. 12 stakeholder interviews across 4 service locations in US and UK. 3 sprints. One mandate: make a dense enterprise system behave predictably without stripping the depth that power users depend on.

The key insight that drove every decision: in enterprise financial systems, clarity and consistency drive operational performance more than feature depth. The redesign surfaced role-aware access transparently, embedded ticket handling directly into contextual views, and simplified payment state logic so users could identify state faster, understand their access, act from the listing view, and share context without exporting.

Post Launch: Launched and awarded in 2025. Ticket actions down to under 3 minutes. Processing time under 4 minutes. Task satisfaction at 75–80%. Ticket volume per team brought under 2,000. A payment blotter that finally worked like a command center.