We build health AI that runs locally on your devices, teach organizations how to deploy it responsibly, audit what's already in production, and partner with institutions on applied research.
A Delaware public benefit corporation.
Powered by the Evidence Cycle: Discover, Design, Build, Test.
Use the app yourself. Train your team to evaluate AI claims. Audit what you've already deployed. Or partner with us on the applied research underneath.
Free iOS app plus a paid Mac platform. Local-first, open-weight health AI on consumer hardware.
Read more ↓Evidence-Based AI for Life Sciences and Healthcare. Public cohorts, enterprise cohorts, executive intensives.
Read more ↓The Health of AI Audit. Quality stability, cost, and continuity for the AI you've deployed — or are considering.
Read more ↓Applied R&D partnerships with government, academic, nonprofit, and corporate institutions.
Read more ↓Local-first health AI on hardware you already own. A free iOS app on TestFlight today, with a paid Mac platform launching late May. Open-weight models run entirely on-device — no cloud, no API keys, no data leaving your device.
Workforce development for teams that need to evaluate AI claims rigorously instead of being captured by vendor narratives. Available as public 8-week cohorts, enterprise programs, and executive intensives. Developed through the Stanford Medicine Health Futurists program and taught at Morehouse College.
An independent evaluation of the AI you've deployed or are considering — built around three questions vendor pitches don't answer.
Cloud-only vendors update, swap, deprecate, and retrain models on their schedule, not yours. The performance you validated last quarter may not be the model you're running today. We document what the model actually is, what the published evidence supports, and what changes would invalidate that evidence — so you know when your AI has quietly become a different product.
Most health systems and life-sciences teams already have the compute they need to run local-first AI on consumer and prosumer hardware. We benchmark your specific use case against the cloud bill you're being quoted, and show where local, hybrid, or cloud is genuinely the right answer for the work.
What happens to your clinical workflows if a frontier API gets export-controlled, repriced, deprecated, or simply throttled? We map your AI dependencies the way operations leaders already map vendor risk for the rest of the supply chain, and identify the parts of your stack that should be on local-first infrastructure for continuity reasons alone.
We partner with government, academic, nonprofit, and corporate institutions to design and fund applied AI research. Engagements run from proposal generation through named research partnership to embedded research entity. Bring us the institution; we bring methodology, technical capability, and on-device AI infrastructure.
We build on open-weight foundations organizations can inspect, run, and own.
Health AI runs on consumer hardware, with no data leaving the device unless the user chooses.
Every claim gets the Evidence Cycle treatment: Discover, Design, Build, Test.
A Delaware public benefit corporation with the public-benefit obligation written into the charter.
A track record of working with leaders across the public and private sector, from NIH and the National Academy of Medicine to Stanford, Morehouse, and MITRE.
Twenty minutes is enough to know if we're a fit. We'll follow up with a scoped plan and next steps — not a sales sequence.
