1 in 4 people worldwide will have a stroke in their lifetime. The first hour decides whether the next thirty years look like recovery — or a second life spent in rehabilitation.
For decades the question has been how quickly can we reach a hospital. We think the better question is how quickly can we recognize what's happening.
The answer is already in your pocket: a camera, a microphone, a neural engine. Only 1 in 5 people make it to the hospital within the first two hours — not because help isn't available, but because nobody recognized the signs in time.
We built something that does.
Stroke is the world's most expensive preventable emergency — and the one category where early detection rewrites the bill, the outcome, and the rest of a life.
No wearables. No new hardware. code:blue uses the existing camera and microphone on any phone, laptop, or webcam — quietly, wherever you are.
Nothing leaves without your permission. If code:blue detects something, you choose who gets notified — a family member, a caregiver, or both.
Models are trained on real stroke footage in partnership with UCSF Health. Every detection model is reviewed by our clinical advisors before deployment.
When code:blue detects a possible stroke, it notifies your trusted circle and contacts emergency services — with your location and condition already shared before help arrives.
Our UCSF pilot (n=30) achieved 100% detection accuracy. Kaiser Permanente reviewed the results and requested a 1,000-patient follow-up study — the next step toward full clinical validation.
Building a product that sits between clinical medicine, consumer software, and the FDA requires a specific, unusual mix of people. We went and got them.
.png){kind=logo}
If you invest in healthcare, early-stage AI, or consumer devices that touch regulated medicine — or if you run a health system thinking about pre-hospital triage — there's a short window to get on the right side of this one.
