← Home
A denied loan. A flagged risk. A process no one could see.
Designing a shared investigation workflow that gave three conflicting roles a single source of truth and the structure to investigate every failure with confidence.
Overview
Data quality isn't an abstract concept at J.P. Morgan Chase. It's the foundation of decisions that determine whether a client gets approved for a loan, offered a higher line of credit, or flagged as a financial risk incorrectly. When a rule flags a failure, someone has to investigate. But the system built to support that investigation was working against the people using it. An enormous volume of information compressed into an interface with no way to distinguish what mattered from what didn't. Three roles who each held a piece of the process, with nothing connecting them and no shared view of the truth. As the UX designer and researcher on this project, I was responsible for the full design lifecycle from discovery through launch. The domain was deeply technical. Understanding what each role actually does required sustained immersion, observing workflows, mapping process stages, and conducting individual interviews and usability tests across all three roles. That groundwork shaped every design decision that followed.
When a data quality rule flags a failure, someone needs to investigate. Was the flag legitimate, or was it triggered by an anomaly like the economic disruption of a global pandemic? The answer determines whether a client gets approved for a line of credit, denied a loan, or flagged as a risk incorrectly.
At J.P. Morgan Chase, that investigation process was entirely manual. No shared workspace. No status tracking. No visibility into who owned what or where a failure stood in the resolution process. Everything lived in email. Rule failures got lost. Resolution timelines stretched.
🚧
The platform wasn't just outdated. It was creating risk in a system designed to manage it.
Solution
The redesign wasn't about modernizing the visual surface. It was about solving a structural accountability problem that no amount of aesthetic polish could fix.
Three solutions mapped directly to three problems:
Bring structure to overwhelming data
Addresses the hierarchy problem. The redesign introduced clear information hierarchy and contextual detail views, giving users access to depth without being buried by it on first glance. Every piece of information was given a place, and a priority.
Design a workflow, not just an interface
Addresses the disconnected roles problem. The redesign mapped the failure resolution process explicitly, giving each role a clear view of where they stood, what was expected of them, and what came next. For the first time, all three roles were working from the same process.
Make transparency a feature, not an assumption
Addresses the visibility problem. Research surfaced the transparency gap. During design review, while restructuring the data hierarchy, it became clear that ownership and status needed to be visible at a glance. View Assignees wasn't something users asked for. It was something they immediately recognized as essential the moment they saw it, eliminating the email dependency that had been burying failures for years.
A design decision worth examining
During design review, while restructuring the data hierarchy, something was missing. There was no way to see who owned a failure. No name. No way to reach out. For a process that lived entirely in email, that gap meant the new system risked recreating the same accountability problem in a cleaner interface.
Every user in the review called it out independently. Before incorporating the feedback, I confirmed with the engineering lead that the data was available. The system already tracked case assignment. Surfacing it was a matter of design decision, not technical lift.
When the iterated design came back with View Assignees, users recognized it immediately as essential. Not a nice-to-have. A missing piece they hadn't thought to ask for because they'd never seen a system that could show it to them.
Results
Evaluated against Nielsen's Usability Heuristics across 11 participants representing all three user roles.
The redesign produced significant improvements across all three dimensions, with the most dramatic gain in aesthetic clarity. Analysts reported significant improvements in site clarity, navigation, and process tracking, giving all three roles the confidence to investigate failures and make decisions without second-guessing the system.
Usability
2.7 → 4.6
out of 5.0
Flexibility & efficiency
2.0 → 4.2
out of 5.0
Aesthetic clarity
2.7 → 4.6
out of 5.0
📣
There's more to this story than what's here. The full designs, process, and results are available under NDA. If you'd like to see the thinking behind the decisions and dig into what shipped, I'd welcome that conversation.
