In the first minutes of an emergency, the person on scene usually isn't a medic. SaveVision streams their point of view to a remote clinician who sees what they see and guides life-saving care — by voice, by drawing on the live image, step by step — until professional help arrives.
In the first minutes of a serious injury, most preventable deaths come from things a bystander could actually treat — catastrophic bleeding, a blocked airway. The knowledge exists; it's just almost never in the room. SaveVision lends it over a wire.
From the first tap to guided care, all in one flow.
From the glasses or a phone, you reach the operations centre. Your live point of view links automatically — no setup, no fiddling under pressure.
A real medic watches your live view, places the scene on a map with the ambulance ETA, and works out what's actually going on.
Voice, drawings on the live image, reference photos, a map on the display — and any AI suggestion is approved by the clinician before it reaches you.
You only ever see what the remote expert chooses to send. Nothing fires at the person on the scene on its own — one-way, by design.
Mark the exact spot to press, or where the tourniquet goes — right on the limb you're looking at.
Push real first-aid images — tourniquet, CPR, recovery position — straight onto the display.
A real map on the glasses with points, routes and the ambulance ETA — switched on when it helps.
A vision model suggests the next step; the clinician approves it before anything is sent.
Callers, the situational map and the task list — kept in sync in a single operator view.
Works on smart glasses with a display, or falls back to any phone camera.
SaveVision is designed for the realities of pre-hospital and humanitarian work: long response times, scarce clinicians, and untrained people who have to act now. It extends the reach of the experts you already have.
Long ambulance ETAs, where a bystander is the only one on scene for twenty minutes or more.
When responders are stretched thin, one clinician can guide several scenes at once instead of one.
Where trained medics are scarce and dangerous to move — guidance travels instead of people.
A force-multiplier for volunteer responders and event medical cover, with a clinician on call.
We'd rather show you something real than quote numbers we haven't earned. Here's exactly where SaveVision stands and where it's going.
Live point-of-view, operator console, drawing, maps and AI-proposed guidance — running today over encrypted Matrix and WebRTC.
Controlled trials with licensed clinicians and a partner organisation — measuring time-to-guidance and decision confidence, with full consent and review.
Hardening for low bandwidth and offline fallback, regulatory and clinical validation, and integration with existing dispatch.
This is medical data in someone's worst moment. We treat it that way — and we'll happily walk your data-protection team through it.
Video and guidance run over Matrix (Olm/Megolm). Not even the server can read what's said.
Run it on your own infrastructure in your own jurisdiction. Casualty data never has to leave your control.
The clinician decides; the bystander acts only on explicit approval. No automation surprises.
Read our data-protection & privacy approach →
No magic. SaveVision is built from open, well-understood pieces — which is exactly why we're comfortable using it somewhere this serious.
Matrix end-to-end encryption • WebRTC live video • self-hosted homeserver • a vision model with a human in the loop
SaveVision is built by engineers and clinicians who kept running into the same story: someone needed help, and the people right next to them didn't know what to do. We're building the thing we wished existed that day.
Aslan had the original idea for SaveVision and builds it end-to-end alongside Martin — the glasses app, the operator console and the realtime pipeline that connects a bystander's camera to a clinician. They split the engineering equally.
Martin builds SaveVision end-to-end alongside Aslan — the same stack, top to bottom — with a sharp eye for how it has to hold up under real pressure, when a panicking bystander and a busy clinician both depend on it.
Practising clinicians keep every guided step grounded in real pre-hospital care — and we're looking for more.
We're looking for clinical advisors, pilot partners in emergency and humanitarian response, and early backers. Tell us a little and we'll get back to you.
Two minutes, no signup. Watch a bystander get talked through an emergency in real time.