Selected work
What I built, and why.
Each project began with a different problem and a different measure of success. Here is what I chose to build, the decisions that shaped it, and where the work landed.
Professional product work
01
Commercial productEnterprise healthcare data platform
Led the evolution of a fragmented healthcare data environment into a self service platform supporting cohort development, commercial planning, targeting, and enterprise client delivery.
- 65+ monthly active users
- 20+ engagements*
- 24 hours to 2 to 3 hours
* Engagements and associated revenue cover the 2024–2025 period.
View the work02
Professional case studyHealthcare data vendor strategy
Evaluated healthcare data partners across coverage, freshness, match rates, usability, cost, and strategic fit to inform a major platform and vendor decision.
- ~25–50% projected data spend reduction
- ~30% usable coverage expansion
Approximate ranges from the vendor evaluation.
See the project work03
Pilots and deployed featuresAI product decisions
A collection of AI product experiments covering knowledge retrieval, analyst routing, summarization, code recommendation, and natural language data access, including features that were deliberately not moved into production.
- 5 to 6 pilot users
- Approximately 50 to 90 queries per week
- ~60% estimated assignment time reduction*
- ~30 to 45 min reduced to 15 to 20 min in supported interpretation workflows
* Routing figure is an estimate from POC testing. Retrieval stayed a pilot; Text to SQL was not advanced to production.
See the project workIndependent product builds
04 · Clinical trials · AI product
Interactive demoClinical Trial Diversity Studio
A redesign of the ClinicalTrials.gov experience with AI assisted discovery and trial diversity workflows.
05 · Automotive aftermarket · AI-assisted commerce
Prototype in developmentAutomart
An agent assisted spare parts discovery and ordering experience for business and consumer users, with a longer term marketplace direction.