Anpheros Platform · for healthcare startups

Build healthcare apps without building the medical data infrastructure.

Use interoperable medical data, developer-friendly APIs and authorized access to build healthcare applications and AI-powered products.

Sign in with Google, press one button, get a sk_test_ key. Your sandbox project comes with its own 30 synthetic patients. No card, no call.

What Anpheros is

Anpheros is the interoperable medical data infrastructure for healthcare applications and AI: a managed HL7 FHIR R4 record per patient, a REST API over the same data, patient consent, provenance and audit, hosted in the European Union. You build the product; the medical record, its standards and its access rules are already there.

Who it is for

Healthcare and digital-health startups, developers of AI health products, teams building patient-facing apps, and integrators connecting labs, clinics and apps.

What it is not

Not a clinic-management system or an EHR. There is no scheduling engine, billing, video visits or clinician charting screen. Anpheros is the data layer underneath your product, and the same record can serve several independent apps.

Where it stands

Private beta. The API, sandbox, SDKs and consent flow work today. Hosting is single-zone and there is no contractual SLA yet. What is and isn't in place

What you no longer build from scratch

Every health product needsOn Anpheros, today
Patient records and a data modelHL7 FHIR R4 record, 26 resource types (Observation, Condition, MedicationStatement, DocumentReference…)
Standard medical codesLOINC, ICD-10 and ATC recognised on write; LOINC and ATC search built in
An API and its documentationREST /v1 and FHIR R4 /fhir/R4, OpenAPI specification, TypeScript and Dart / Flutter SDKs
Patient-authorized accessOAuth 2.1 with PKCE and SMART on FHIR scopes; consent with expiry and revocation, visible to the patient
Audit and provenanceEvery read and write logged; every version kept; who wrote each record
Lab data intakeCSV, HL7 v2 ORU^R01 and JSON reports → laboratory Observations
Context for AIA context API that returns only the relevant parts of a record, within a token budget, each item labelled with its source
EventsSigned, retried webhooks
Test data30 synthetic patients per sandbox project, six months of history, resettable

What startups build on it

Each use case below links to a runnable starter project or tutorial. All of them run against the free sandbox with synthetic data.

Works today · starter project

AI health assistant

Ask a question about a patient's record: Anpheros builds a labelled context within a token budget, and your server sends it to the model you choose. Credentials never reach the model, and the answer is not a diagnosis.

ai-context-llm →
Works today · tutorial + starter

Patient health app

A Flutter app that reads and records measurements through the Dart SDK, with the person signing in to share their own record. Anpheros' own app, Anpheros Daily, keeps its records on the same platform.

Flutter tutorial →
Works today · starter project

Laboratory integration

Send a whole lab report as HL7 v2 or CSV; get FHIR laboratory Observations with LOINC codes and reference ranges. Markers it cannot map are listed, and re-sending a report never duplicates results.

lab-integration-ts →
Works today · starter project

App backend on FHIR resources

A TypeScript backend that keeps the key server-side, lists patients, summarises a record, writes a validated measurement and turns API errors into answers your frontend can show.

backend-ts →
Works today · starter project

Authorized access to a patient's record

A web app asks a person for consent (OAuth 2.1 + PKCE, SMART on FHIR scopes) and reads only what they allowed. In production, real records are reachable only this way, or for patients your own project created.

consent-app-ts →
Works today · MCP server

AI agents with tool access

Connect Claude or another MCP client to your sandbox: list patients, read FHIR resources, build context and import lab reports, as tools with the access you allow.

MCP guide →

From zero to your first API call

  1. Get a sandbox key. Sign in on the dashboard and press Get a sandbox key. The key is shown once; keep it on your server.
  2. Make a request. Your project already has 30 synthetic patients:
    curl https://platform.anpheros.com/v1/patients \
      -H "Authorization: Bearer $ANPHEROS_KEY"
  3. Install an SDK.
    npm install @anpheros/sdk        # TypeScript / Node
    flutter pub add anpheros_sdk     # Dart / Flutter
  4. Run a starter project: write and read a record end to end.
    git clone https://github.com/anpheros/anpheros-sdk
    cd anpheros-sdk/examples/quickstart-ts && npm install
    ANPHEROS_KEY=sk_test_… npm start
  5. Test consent with consent-app-ts: the sandbox consent page is marked TEST.
  6. Go to production. Same code, a sk_live_ key, after your organisation is verified and signs a data processing agreement.

How access and data are protected

Sandbox and production are separate

Sandbox keys reach a separate database with synthetic patients only. A prototype, including one written with AI coding tools, cannot touch real records.

The patient decides

Apps read a person's record only with their consent, limited to the types of data they chose and to a fixed duration; they can revoke it at any time and see every access.

Every write has an author

Provenance records who wrote each resource (patient, device, practitioner, import), and earlier versions stay in the history.

Your data stays portable

Records are standard FHIR R4: Patient/$everything exports a whole record and $summary an International Patient Summary. Hosted in the EU.

Anpheros does not claim medical-device certification, ISO 27001, SOC 2 or a contractual SLA. Read the security model and the reliability page for what is in place today.

What it costs

Free

Sandbox

The full API with synthetic data, no time limit: 30 synthetic patients plus up to 500 of your own per project, 10,000 writes a day, 250 MB of documents, up to 5 sandbox projects.

From €49 / month

Production

Monthly plans by stored patients: Starter €49 (2,500), Growth €199 (10,000), Scale €799 (50,000), prices excluding VAT. The first month of production is free. Production access is opened for each organisation after verification and a signed data processing agreement.

Start in the sandbox today

A key in one click, synthetic patients included. When you are ready for real users, write to us and we set up production with you.