Say “Hey Cody.”
Keep your hands
on the patient.
A voice-first clinical co-pilot for EMS. Hands-free, protocol-cited answers in the moment — and the run report and QA review drafted when the call ends.
A contraindication caught mid-call.
“Hey Cody, I’d like to give nitro.”
“Heads up on nitro. Last blood pressure was 88 over 60, below the 90 systolic threshold per Maryland protocol. Nitroglycerin is contraindicated — the protocol lists a fluid bolus for hypotension.”
The patient’s pressure had dropped. The protocol threshold was crossed. COD3 surfaced both against the live patient record — in the moment, with the citation, while the medic’s hands stayed on the patient. The medic decides. Cody makes sure the decision is informed.
Two jobs, one medic, no backup.
Documentation eats the shift.
Agencies report about 15 minutes to hand-write a single narrative, and 60% of EMS leaders name documentation a top challenge (EMS1, 2025). It happens after the handoff, before the next call.
Pediatric doses go wrong under pressure.
Published Michigan EMS data put pediatric dosing errors near 35%, and a statewide paper dosing reference didn't move it: 34.7% before, 35.2% after (Prehospital Emergency Care, 2012 and 2024).
Retrospective QA, mostly by hand.
Quality review usually happens days or weeks later, with reviewers reading narratives back to spot protocol gaps. Feedback reaches the medic long after the call it was about.
Two products in one workflow.
COD3 sits with the medic during the call and with the QA team after it. Same conversation, two outputs.
In the moment
- Voice-first. Hands-free. "Hey Cody, dose epi for a 22kg child."
- Pediatric doses computed in code, not by the AI. Every answer cites the protocol page.
- Contraindications surfaced from what's been said, in stating language. The medic makes every decision.
- "Hey Cody, epi's in" logs the code event to the second, with a tone. A code clock tracks CPR, epi and shocks.
After the call
- A PCR narrative draft built from the call: only what was said, nothing invented.
- A QA review of every call, with protocol-cited flags in seconds, not weeks.
- A black box: every answer Cody gave is saved with the call, with its citation and source.
- Trend views so an agency can see patterns across providers and runs.
How it works.
Speak.
Medic talks to the patient and the team. COD3 listens passively, and a pause button stops it. The wake word — Hey Cody — or a push-to-talk clicker opens a question about dosing, protocols, calculations.
Capture.
Vitals, interventions, narrative beats — extracted live as the call unfolds. Code events are time-stamped to the second. The medic sees what's been captured, on a glanceable card.
Review.
After the call, COD3 drafts the PCR narrative and a QA review against the protocols. The medic reviews and edits, then copies it into the agency's ePCR.
Improve.
Trend dashboards show the QA team where the gaps are, and each medic sees their own call reviews while the call is still fresh.
The clinic got AI. The ambulance is different.
Ambient AI’s breakout tools were built for a clinician sitting at a workstation. Those assumptions don’t hold in the back of a moving ambulance.
Hands on the patient, not a keyboard.
A clinic scribe assumes someone at a desk who can type and read a screen. A medic running a code can do neither. Voice-first, hands-free isn't a feature here — it's the only interaction that works. A clicker and an earpiece keep it discreet in front of the patient.
The decision happens in the moment.
In-clinic AI summarizes a visit afterward. Prehospital care is a series of time-critical calls made before the patient reaches a hospital. The value is help during the call — the documentation comes with it.
Built by a paramedic, for the field.
COD3 is designed around how EMS actually runs a call — protocol-grounded, citation-backed, and deliberately non-directive so the medic stays the decision-maker. Not a hospital tool adapted down to the ambulance.
Why now.
Ambient AI proved itself in the clinic, and AI charting has now reached EMS, often bundled free. The in-call, hands-free clinical layer is still unclaimed at scale.
spent on ambient AI scribes in 2025.
Up 2.4× year over year (Menlo Ventures).
Abridge's valuation, June 2025.
Ambient documentation's category leader (Axios).
of 2025 US digital-health venture dollars went to AI startups.
Rock Health, 2025 year-end funding report.
“Ambient AI reached the ambulance as a scribe. COD3 adds the safety layer in the moment.”
We’re not another charting tool.
EMS charting software now takes dictation and drafts narratives with AI, often at no added cost. Protocol and dosing apps are built tap-first. COD3 works alongside the medic hands-free through the whole call — the documentation and QA come with it.
Get in early.
Design-partner agencies shape the product: their crews, their protocols and their QA priorities come first. If your agency wants to be one of the first three, write us.