Case Study

Thomson Reuters

Fraud Detect utilizes expert machine-learning technology for government programs in Healthcare, SNAP, and Unemployment.

The Challenge

The existing application resembled a website more than a web application. The UI had numerous inconsistencies, from varying font families and sizes to non-brand colors and insufficient contrast, which failed accessibility standards. Additionally, there were various user experience challenges, including confusing UI elements and the absence of a dashboard to provide users with actionable leads.

The Solution

I traveled across the country to meet with clients, observing firsthand how they used—and didn’t use—the application. From these insights, I developed user personas and documented pain points and opportunities for improvement to guide our team’s development efforts.

Throughout the design process, I collaborated closely with SMEs, BAs, developers, executives, and clients. By building interactive prototypes, we could test interactions and gather valuable user feedback. This approach led to a smooth rollout, with clients feeling engaged in the redesign process and excited for the much-needed application update.

The Result

Fraud Detect identified $50 million in overpayments, resulting in $18.7 million in recoveries by the Medicaid Inspector General. The system was deployed to thousands of government professionals across various state and federal agencies nationwide. The new design and user experience also played a key role in securing millions of dollars in government contracts across the country.