"Save As" feature for an AI-enabled fraud-detection application cemented a lucrative state contract

21CT is a maker of fraud detection tools for the state of Texas. Their product, Torch, was already in use but lacked a key "save as" feature. With many complexities, we built requirements, designed, prototyped and implemented this feature based on my work with the 21CT team of stakeholders and developers.

The problem

In this project, I was tasked with imagining and prototyping a ‘Save’ feature into a 21CT product. In theory it’s a basic function and an easy win. In practice, the application functioned more like a web browser than a word processor, and the meaning of ‘save a file’ meant different things depending on what kind of entity was going to be saved.

The world of medical fraud detection, there were many entities that could be saved. Individual providers, clinics, bank accounts and other objects were items that could be flagged for later retrieval. Entire networks, or parts of networks were also of interest to investigators.

The ability to save also meant the ability to categorize and create whole file structures. All of this needed to be in place before “save as” would be useful to the end user.

I mapped the data structure of the new feature.

Interviews with fraud investigators for the state of Texas  helped clarify user needs and requirements.

We paid a visit to the Austin office of the Texas Inspector General. The fraud investigators there walked us through a day in the life of a fraud investigation. They explained how they were using the product currently, as well as things they needed to make it truly a part of their daily routine.

We built a clickable, testable prototype with 60+ screens to test with end users.

After the prototype had been socialized, I used video to  communicate requirements asynchronously to a distributed team.

With such a complex flow, I wanted to make sure that documentation of interactions especially were over-communicated to the team of developers and stakeholders.

I created a style guide to give the developers a detailed reference for use during the sprint development cycle.

Column guidelines helped define common layout schemes for the application.

A UI style guide helped developers standardize the look and feel—and helped the UI developers move more quickly during implementation.

The feature was released with great success to fraud investigators working across multiple U.S. jurisdictions.