Thrive AI Health

Introduction

The Thrive AI Health Partner APIs connect partner applications to Thrive AI Health's platform; wearable integrations, personalized coaching, activity and sleep tracking, lab analysis, and AI-generated health insights.

The API is REST over HTTPS for synchronous calls and partner-owned message queues for asynchronous event delivery. An official React Native SDK is also available; contact your Thrive AI Health onboarding contact for the package URL and access credentials.

Looking for a full endpoint list? Jump to the Route index.

Getting started

Receive credentials

Your Thrive AI Health onboarding contact provisions two secrets per environment, a tenant API key (used in the taih-tenant-key header) and an event-signing HMAC secret (used to verify events delivered to your queue), along with the base URL of the sandbox environment for integration testing.

Store every tenant API key and HMAC secret in a backend secret manager; neither is safe to embed in mobile or browser clients. Sandbox and production credentials are distinct; don't reuse sandbox values against production or vice versa.

Create your first member

Call POST /v1/users with the tenant key to register the partner-scoped user_id you will use in every member-scoped call.

Mint a member token

From your backend, exchange the user_id for a short-lived member bearer token via POST /v1/auth/token. Forward the token to your client.

Make a user-scoped call

The client calls /v1/users/{user_id}/… with authorization: Bearer <access_token>. A good first call is GET /v1/users/{user_id}/health-profile to verify end-to-end wiring.

Register an event queue

For event streams you consume (lab results, daily device data, coaching nudges), register a destination queue via POST /v1/admin/org/queues/{queue_name} on the admin surface.

Versioning

The API version is carried in the URL path (/v1/…). Only v1 is currently published. Every change inside v1 is additive and backwards-compatible; breaking changes ship only in a new major version.

Deprecation policy

When a new major version is published, the previous version enters a predictable wind-down:

  1. Announcement: partners receive email and changelog notice at least 6 months before deprecation takes effect.
  2. Maintenance mode: the deprecated version receives only security patches and critical bug fixes for at least 12 months after the announcement.
  3. Sunset: the deprecated version is retired 18 months after the announcement.

Partners should rehearse migrations against the sandbox environment before cutting production traffic.

Support and resources

Interactive reference

Browse, search, and try endpoints against your sandbox credentials.

Route index

Every endpoint grouped by domain.

OpenAPI spec

The URL is provided by your Thrive AI Health onboarding contact.

Support

Operational questions, incident reports, and integration help go through your Thrive AI Health onboarding contact.

On this page