The hardest problem in health isn't the science. It's getting the science to people.

Where I work now, as CEO. We measure a six-analyte inflammatory panel from a blood sample collected at home, reported as individual values alongside a normalized composite. The underlying science is published: an IL-18-centered network described in PLOS ONE (2020), and a 2,201-adult Framingham Offspring cohort examined in Stroke (2024). That work came out of our founder's academic lab and is licensed to the company. Supported by an NIH SBIR grant (R43NS139780), run in CLIA-certified labs. Consumer access opens in stages, and we measure the same panel as an endpoint for research teams.

Co-founder. A decentralized approach to health data: self-sovereign ownership, gamified health quests, and protocol tools for creators. We published a light paper in March 2022; the two demos below show the consumer challenge app and the creator protocol dashboard. We wound it down in 2022: the protocol worked as designed, but there was no realistic base of users who wanted that level of custody over their own health data.

CONSUMER CHALLENGE APP DEMO, 2022 · 0:44

CREATOR PROTOCOL DASHBOARD DEMO, 2022 · 0:51

SYSTEM DESIGN, COLLAB HEALTH, 2021

Seven years as CEO, building a personalized health platform: monitoring, recommendations, and e-commerce, built on data from more than 400 devices and sources. We were doing machine learning on personal health data when that still meant building the models yourself. We went through 500 Startups (Batch 20) and raised more than $2M; it wound down in 2021, after a funding round collapsed during COVID.

Personalized nootropics. I ran research and development: formulations, and the protocols for measuring whether they actually worked, from a kitchen lab through clinical studies to a shipped product. I left in 2014 to start Biomarker Labs, with the aim of bringing validation to an industry rife with snake oil.