Overview
Integrating with Longevity AI gives your team a direct, secure connection between your existing system and Florence, Longevity AI's clinical intelligence platform. Patient data flows in, Florence's clinical outputs flow back, and your clinicians work in the systems they already use.
To make integration fast and predictable, we follow a simple three-step process: sandbox validation, API enablement, and production cutover. The same path works whether you're a health system integrating into an enterprise EMR, a multi-clinic group connecting an internal patient management system, or a clinical platform embedding Florence into your own product.
Step 1: Provide a sandbox environment
We always begin in a non-production environment so the connection can be developed and validated safely, with no risk to live patient data or clinical workflows.
Your sandbox should include:
- A sandbox that mirrors production, with the same data structure, fields, and custom configurations, so behavior is consistent across environments
- A super-user account (for example, doctor-level access) with full permissions for the duration of the build
- The ability to create members and upload tests or files
- Access to the system's API
A note on permissions during sandbox: the user account we work with should have full permissions during the build to avoid delays. Access can be tightened in production once the integration is validated.
Step 2: Enable API access
Once the sandbox is ready, we connect to your system through its API to exchange the data Florence needs.
Required API capabilities:
- GET and POST access to the resources we'll work with: Members, Doctors, and Tests or Files
- Secure authentication (OAuth or token-based), including:
Recommended (optional but high-value):
- Webhooks for real-time updates, so Florence stays in sync the moment a new member is created or a new test is uploaded
Real-time updates substantially improve the clinical value of the integration. Without webhooks, Florence relies on polling, which introduces lag between data arriving in your system and Florence acting on it.
Step 3: Go live in production
Once the integration is fully tested in sandbox, the same setup is replicated in production. No additional development is required, just an environment switch.
Your production environment should include:
- A production environment with structure and configurations similar to sandbox
- A super-user account with the required permissions, scoped to the integration's actual needs
- API access and authentication credentials
Ongoing collaboration
A smooth integration depends on two things on your side:
- A technical point of contact who can answer questions, manage permissions, and help resolve issues
- Availability to quickly resolve access or configuration questions as they come up during the build
Most issues that slow integrations down are access or permissions related, not technical. A responsive point of contact resolves them in hours instead of days.
