Composable
Call the whole care cycle or a single module. One request shape across every endpoint.
Developers
Five REST endpoints that return structured, scored JSON, with clinician review built in, not bolted on.
Endpoints
The whole cycle or one module. Same request shape, same review semantics.
/v1/origin/assessmentOriginDrop adaptive pre-visit screening into your own intake flow.
/v1/trace/noteTraceTurn a transcript into a structured, reviewable draft note.
/v1/bridge/scheduleBridgeAutomate between-visit check-ins.
/v1/lens/trajectoryLensShow longitudinal trajectory across instruments.
/v1/signal/evaluateSignalSurface patterns that may need attention.
# surface patterns over a 14-day window # base URL is issued with your API key curl $EMP_API_BASE/v1/signal/evaluate \ -H "Authorization: Bearer $EMP_KEY" \ -H "Content-Type: application/json" \ -d '{ "patient_id": "pat_8fa21", "window": "14d", "inputs": ["check_ins","journal"] }'
{
"review_priority": { "level": "high", "score": 0.71 },
"indicators": ["sleep_decline", "isolation"],
"route": "care_team",
"confidence": 0.93,
"review_required": true
}Why build on Empaithy
The hard parts handled, so you ship a product, not a research project.
Call the whole care cycle or a single module. One request shape across every endpoint.
Bearer-token auth, encrypted transport, scoped keys, and full request audit trails.
Clinical reasoning across languages, with no pre-translation step.
Each structured field carries a score and provenance, so you know what to route for review.
Quickstart
Receive a project key scoped to your environment, with a sandbox to build against.
Post a transcript, note, assessment response, or signal window to the relevant module endpoint.
Get back scored JSON with per-field confidence and a clear flag for anything that needs review.
Surface outputs in your product with review controls, never as autonomous decisions.
// npm i @empaithy/sdk import Empaithy from "@empaithy/sdk"; const emp = new Empaithy(process.env.EMP_KEY); const r = await emp.signal.evaluate({ patient_id: "pat_8fa21", window: "14d" }); console.log(r.review_priority.level); // "high"
# pip install empaithy from empaithy import Empaithy emp = Empaithy(os.environ["EMP_KEY"]) r = emp.signal.evaluate( patient_id="pat_8fa21", window="14d", ) print(r.review_priority.level) # "high"
Webhooks
Check-in responses and highlighted patterns arrive as signed webhooks, with the same provenance and review state as the REST response.
A patient answers a Bridge check-in, or Signal highlights a pattern for review.
Each request carries a signature header, so you can verify it came from Empaithy.
The event arrives as JSON. Reply with a 2xx to acknowledge.
Failed deliveries are retried with backoff, so build your handler to accept a repeat.
{
"event": "signal.review_item.created",
"patient_id": "pat_8fa21",
"review_priority": { "level": "high", "score": 0.71 },
"drew_on": ["check_ins", "session_notes"],
"review_state": "open",
// cannot close without a documented reason
"requires_reason": true
}Tell us what you are building and we will get you a sandbox key.