Ray, 71
He Called It Nothing
A man on his bedroom floor. His daughter two hours away. You can hear his lungs from the doorway, and he does not want your help.
“I’m fine. You didn’t need to come.”
Text-based AI communication simulation for paramedic services, base hospitals & college paramedicine programs.
Practice What You’d Actually Say
Typed conversations with AI patients who respond to how you speak — not voice, not VR. Built by a practicing Ontario paramedic. Version 1 is built, and the first pilots open Summer 2026.
One typed line at a time, and the call shifts. Runs in your browser, no signup.
The conversation decides the call.
Not the protocol. The minute before it.
The Scenario Library
The calls that don’t fit a protocol.
Each one targets a call type research flags as among the hardest to communicate through: care refusal, acute distress, and mental-health crisis. Scenario 01 is live now; 02 and 03 are in development for the first pilots.
Cass, 19
Just Take Me to Him
Blood is running down her face and she won’t let you near it. She’s nineteen, furious, terrified, and the only thing she’ll talk about is her boyfriend. Getting her to let you actually assess her is the hard part.
“Stop. I don’t care about my head. Where is he?”
Daniel, 15
He Already Had a Plan
A teenage boy in crisis, with injuries he’s caused himself and a plan he’s already made. The clinical care is the part you know. Keeping him talking, and keeping him here, is the part nothing trained you for.
“You can’t fix this. Nobody can.”
Inside a Scenario
The character responds to how you speak, not what you say.
There is no right phrase to find. Trust, agitation, and engagement shift quietly beneath every exchange. Rush, and the patient closes. Slow down, and the door opens. Learners can’t optimize a number they never see. They have to read the person.
Illustrative animation: a simulated text conversation between a paramedic and Ray, a 71-year-old patient who initially refuses help. As the paramedic slows down and listens, hidden trust, agitation, and engagement meters shift, and Ray agrees to be assessed.
How It Works
Practice the conversation. Then sit with what happened.
Text-based exchanges with AI-driven characters that respond to tone, followed by structured reflection. No pass. No fail. No scripts.
01
Enter the scenario
A brief, field-realistic setup: a call type, a character, a situation. The learner reads it, then starts the conversation cold — a few quiet minutes, on any phone or laptop.
02
Read the person, not the metric
Hidden state (trust, agitation, engagement) shifts on tone and pacing. The hard part of these calls was never the words leaving your mouth — it’s the read and the choice. Typed practice trains exactly that decision, again and again.
03
The call ends like calls end
Stabilization, shutdown, resistance, or connection. No score is tallied. Every outcome is a real outcome the learner has to own.
04
Reflect — and keep one anchor
Structured prompts turn the call into a habit, solo or as a facilitated group debrief with the included guide. Every scenario ladders back to one word — PRESENCE™ — a medic can actually recall mid-call.
A Different Category
Most simulation scores the interview. We train the moment it can’t.
Clinical simulators are built to measure reasoning: history, vitals, the right differential. Empathia is built for the human encounter that decides whether any of that gets to happen.
Conventional clinical simulation
Empathia Learn
What it trains
ConventionalClinical reasoning, history-taking, documentation
Empathia LearnThe emotional encounter, the minute before the protocol
How the patient behaves
ConventionalFollows a clinical script toward the correct path
Empathia LearnResponds to your tone, driven by a hidden emotional state
Feedback
ConventionalScored: pass, fail, a competency number
Empathia LearnStructured reflection, never a score
How it ends
ConventionalA correct answer
Empathia LearnA real consequence: connection, refusal, or shutdown
What the learner leaves with
ConventionalA grade in a gradebook
Empathia LearnPRESENCE, one anchor they can reach for mid-call
Built by
ConventionalA software vendor
Empathia LearnA practicing paramedic and researcher
For the People Who Run Training
Accountability without a scoreboard.
Coordinators see participation and coverage, never the conversation or a grade. A learner’s words stay theirs — what you get is proof your people are practicing.
24
Active learners
312
Scenarios practiced
287
Reflections completed
13
Avg. sessions / learner
Learner participation
LearnerScenariosReflectionsCall types practiced
A. Okafor1615
J. Lindqvist1414
M. Whitefeather1311
D. Santos1110
R. Bélanger98
Scenario coverage
Care refusalfully covered
Mental-health crisis88%
Impaired driver / MVC71%
Pediatric & caregiver54%
Agitation & de-escalation44%
Participation & coverage.Export for base hospital audit ↓
Canadian data storage
No PHI stored
Privacy-law aligned (PHIPA)
Never used for discipline — contractual
Everything your privacy and security reviewers will ask, answered up front: Security & privacy →
Who It’s For
Built for prehospital care, start to finish.
For service chiefs
Paramedic Services
Slots into the CME and in-service hours you already mandate, with exportable proof of practice for the calls that draw complaints.
For base hospitals
Base Hospitals & Medical Oversight
Audit-ready participation and coverage records for the competencies hardest to evidence — never surveillance of a medic’s words.
For educators
College Paramedicine Programs
Repeatable practice before placement, mapped to the NOCP → CPCF communication competencies your program is measured against.
Full curriculum-alignment and records detail: Alignment & records →
Built in the field.
Empathia Learn was founded by Sandra Trejo, PCP, a practicing Primary Care Paramedic and researcher in Ontario. She is building the program from both sides of the work at once: the clinician who has stood in those doorways, and the researcher proving what actually helps. Every scenario here comes from a real shift.
Evidence Status
An evidence program, stated plainly.
Nobody in this space has outcome trials yet — including us. Our evidence program is young and we say so: here is exactly where each piece stands.
Manuscript in preparation
When the Protocol Ends: an emotionally adaptive AI simulation platform for communication training in prehospital care
Survey in distribution
Needs-assessment survey with practicing Ontario paramedics
Pilot planned · 2026 to 2027
Prospective pilot study: emotionally adaptive simulation in the field
Frameworks
EARS™ — the listening framework taught inside the scenarios — and PRESENCE™, the one-word anchor: described in the manuscript in preparation; neither has been peer reviewed yet
Getting Started
What a pilot looks like.
One cohort, your real training calendar, a fixed cohort-sized quote — a purchase-order line, not a capital request. Nothing renews until you’ve seen the data and decided.
1
We set it up with you
Onboard your cohort
We create your account and load your roster. No IT ticket, no hardware. Learners get a link and sign in from any device.
2
Where training already lives
Assign it as required training
Runs inside the CME and in-service hours your service mandates, or a course in a college program — facilitated or self-paced.
3
Proof at review time
Decide with evidence
Watch participation and coverage build in the coordinator view, then decide on a full license with exportable records in hand.
Bring this to your service or program.
Version 1 is built, and the first pilots open Summer 2026. Tell us about your team — we reply personally, usually within two business days.
Prefer email? hello@empathialearn.com