This project was a complete redesign of the US Courts probation officer training management application. The existing application was adopted by less than 12% of intended users. The primary goal was to dramatically increase adoption, while also expanding the potential user base beyond probation officers.
I led the design effort and wrote all the pre-logic front-end code, largely designing "in the browser" so that the static prototype essentially became our working application.
The greatest challenge was that each federal court district — there are 94 — has free reign to use whatever technology they wish; there is no centrally mandated administration due to the constitutional nature of the federal judiciary. Consequently, most districts were not using the existing training management tools — which they saw as largely unusable — and had instead adopted third-party or home-grown software...or just continued to use spreadsheets, email, and pen-and-paper. In order to meet our goals we aimed to both radically improve the user experience and expand the utility of the application to handle training beyond just that taken by probation officers.
Tackling semantics was crucial early on. Because of the many solutions already in use for training management, we found that many different terms were being used in different ways. We had to be careful to ensure that the single set of terms that we employed were both meaningful to users and were not confusing given their past usage. Surveys and interviews provided us a good starting point and we then worked through various scenarios, both from a user's and from a technical perspective to ensure that our terminology was the best possible.
Another big challenge was halfway through the project when the limitations of our chosen UI framework — Zurb's Foundation — began to put us into a corner. Our own modifications to the framework began to conflict in numerous ways and I made a daring pitch to the team: to switch frameworks and completely renovate the style of the site at the same time. I had already been experimenting with solutions to our problem and had decided that using a light-weight framework based on Google's Material Design provided us the most coherent design without locking us into another corner. At the same time, the patterns it provided were an excellent foundation for building my own additional patterns that would solve our growing needs. I pitched the pivot and was given the green light. A week later the entire application was refactored and I had created everything we needed to continue our work.