Eight case studies, honestly labeled by what shipped to users vs. what remains a prototype, a discovery, or a decision to stop.
A healthcare purchasing analytics team prices supplier quotes for hospitals. Before any of that judgment could happen, analysts retyped every line by hand. I led the AI + business-rules engine that replaced the retyping. The judgment stays with the analyst.
A B2B procurement checkout with drop-off spread evenly across the flow, not concentrated at one step. That finding ruled out a single fix: iterative A/B testing across several targeted changes, not a redesign.
A solo rebuild of a 20-year-old desktop platform as a modern, AI-first web application. Roughly 60–70% of the legacy system is rebuilt completely; the rest is documented as a prioritized migration roadmap instead of pretended parity.
We built a continuous monitoring engine that uses HTML scraping and Claude 3.5 Sonnet to translate regulatory changes into a strict Obligation → Policy → Action workflow, drastically reducing manual review time.
Researchers review dozens of clinical trials to support treatment decisions. The bottleneck isn't search; it's standardizing buried data like participant counts, study designs, and conflicts of interest. We built a POC to automate the extraction while enforcing strict human oversight.
0→1 radiology platform, multi-tenant, FDA SaMD-compliant. We pivoted the MVP from speed-first to interpretability-first, and adoption followed.
A well-scoped, technically exciting AI initiative. Compliance blockers we'd have hit at integration. A recommendation to terminate, before a single sprint was burned.
A single-file, offline-first workspace that unifies tasks, follow-ups, and links across a multi-product portfolio, built because no enterprise tool treats "who am I waiting on" as a first-class thing.
Available for AI product leadership roles, architecture reviews, and 0→1 enterprise execution.