Work

Founded AI document review startup for regulated industries. 40+ user interviews. Chose RAG over fine-tuning for built-in citations. Targeted law students as the entry point into institutional adoption. Review time dropped 71%. NVIDIA Inception accepted.

Building the systems that keep ML models alive after deployment — pipelines ingesting millions of records daily, monitoring that catches drift before it reaches patients, infrastructure auditable under HIPAA. 150M+ patients served.

Built VantaStat to collapse the timeline from orthopaedic injury to specialist consultation from days to minutes. Shipped analytics dashboards, HIPAA-compliant backend, and workflow automation saving ~5 hours of admin time per week per practice.

Built the event pipeline, analytics dashboards, and experimentation infrastructure for a growth platform. Events process in under a second. Dashboard queries that took 12 seconds now take 200ms. The team stopped debating what happened and started debating what to do about it.