First Request

Your first request should confirm three things at once: credentials are valid, the target workload category is recognised, and the response includes enough routing detail for the client to act on it.

Prepare the request

Choose a representative workload category, attach the environment credential, and provide the request payload expected by that category. Start with a non-production environment until your routing policy is validated.

Submit traffic

Send the request to the public API and capture the full response body. The response should include the selected provider, request outcome, and identifiers that let you inspect the decision later in telemetry and logs.

Inspect the decision

Confirm which provider was chosen and whether the selection matched the expected routing policy. If the chosen provider differs from expectation, compare live health and benchmark rank to understand why the routing layer made that tradeoff.

Validate the feedback loop

After the request completes, check that the request is visible in telemetry and that the observed latency or error status was recorded. This verifies that runtime behaviour is feeding back into the platform.

Next step

Use Telemetry Overview to review what operational signals are captured after the call completes.