Designing a more efficient process and tool (2025)
Integral leverages technology to help companies make their Real World Data HIPAA compliant, empowering users to maximize data utility and patient privacy. I was the first full-time product designer for the company and owned the entire design process from conceptualization to final implementation, working closely with a two-person engineering team and the CTO.
One of my proudest efforts was tackling the data dictionary reconciliation process. Take a look at the mess we were in regarding the mismatch between actual data and data dictionaries:
I had a mission: make the data dictionary mapping process easier, so we could get to the expert certification process faster and less painfully.
I had to facilitate alignment on several key decisions. Two that guided the design of the process and tool were:
The connected data would be the source of truth.
Users would be given the agency to validate their data dictionaries.
As a team, we agreed that with these foundational decisions made, we could surface connected data to users and bring them into the formerly opaque process to efficiently and transparently map descriptions to data. But of course they wouldn’t be alone! We leveraged AI to help in two places. AI would:
Help match connected data to uploaded data dictionaries
Give a first draft description to data variables without matching descriptions
After several rounds of iteration and user testing, here’s a peek into where we landed:
We built a brand new process and tool that enabled users to reconcile their own data with their own data dictionaries, freeing up them and our team to focus on more important work. All while:
Using as many existing design system components and patterns as possible
Maintaining consistency and flexibility with other steps of the expert certification process
Growing user trust in a software solution that didn’t exist anywhere else in the industry