Going to production

A checklist for moving your Vantage integration from sandbox to production safely.

Sandbox and production are separate environments with separate access: your credentials are enabled for sandbox first, and production is enabled when you are ready to launch (via your Terra contact). Before flipping your integration to https://vantage.tryterra.co, walk this checklist.

Access and configuration

• Production access enabled - confirm with your Terra contact that your account is enabled for production ordering. Sandbox access alone does not open production.

• Production webhook URL registered - sandbox and production hold separate webhook URLs. Register the production one: PATCH /api/v1/clients/webhook-url against vantage.tryterra.co, and confirm with GET /api/v1/clients/webhook-url.

• Base URL switched everywhere - audit for hard-coded vantage-sandbox.tryterra.co references.

Integration behavior

• Signature verification passes in production - same signing secret and scheme as sandbox; confirm your first production webhook verifies.

• No reliance on sandbox-only behavior:

  • The simulate endpoint returns 403 in production - make sure no product code path calls it.

  • supplier_item_id is absent from production webhook payloads - if your activation UX needs the kit ID, read it from GET /api/v1/orders/{orderID} (items[].supplier_item_id) instead.

  • Sandbox auto-generates results on a timer; production results take real-world time (days). Make sure nothing times out waiting.

• Duplicate-order protection - production orders cost real money; ensure ambiguous failures are reconciled before retrying (Best practices).

• Failure paths built - results.sample_rejected and results.lab_processing_error need a real user-facing flow (replacement kit, support contact) in production.

Compliance

• Acknowledgment flow live and user-triggered - results must be acknowledged by an explicit end-user action before patients can access them; unacknowledged results shift liability to you and escalations have an acknowledgment_due_by deadline (Acknowledging Results).

• Escalation handling - a plan (and a monitored alerting path in your systems) for results.escalation_raised events, including out-of-hours.

Operations

• Monitoring wired up - watch GET /api/v1/overview (webhook failures, missing results) and check GET /api/v1/webhook-deliveries?outcome=failed when something looks off (Monitoring and debugging).

• Reconciliation job scheduled - a periodic sweep of GET /api/v1/orders?missing=true and unacknowledged ready results (Best practices).

• Support runbook - your team knows to quote order_id / order_item_id / event_id (all strings) when contacting Terra support.