Learning center

Send your first governed AI request

Follow the shortest safe path from approved model supply to a bounded route, hard budget, one-time workload identity, test request, and reason-coded evidence.

For First-time administrators, individuals, small teams, and implementation partners · Updated August 10, 2026
Working product walkthrough

Watch, listen, or read.

The screens come from the sanitized working sandbox. No customer data, credentials, prompt bodies, or provider responses appear in the recording.

  • English captions and transcript
  • No autoplay or third-party video tracker
  • Reviewed 2026-08-10

Transcript: send your first governed request

This walkthrough shows the shortest safe path to a first request through AI Gateway HQ. You will connect approved model supply, create a route, set a hard budget, issue a workload key, and verify the evidence.

1. Connect approved model supply

Begin on Providers. Add a supported provider credential you control, or activate eligible prepaid managed Bedrock. The console stores the secret through the encrypted provider boundary and does not display it back after submission.

2. Give the workload a stable route

Create a route with a clear workload name, then select only the approved provider accounts and models that may serve it. Priority and fallback never override capability, health, policy, or budget eligibility.

3. Set the stop before traffic

Add a hard budget before you generate a key. Choose a limit that matches the owner and environment. Strict enforcement reserves conservative exposure before dispatch, so exhausted spend stops before another provider call.

4. Issue one workload identity

Issue a workload key for this application and environment. Copy the key once into the application's secret store. Do not put it in source code, screenshots, tickets, or shared documents. Revoke and replace it if its boundary changes.

5. Send a small test request

Send a small non-sensitive test to the gateway endpoint using the workload key and route alias. The gateway checks identity, policy, budget, target eligibility, and provider health before it forwards anything.

Verify the reason-coded result

Open Audit and requests, then find the matching request identifier. Confirm which route and provider were selected, the policy decision, upstream attempts, latency, usage, and cost basis. Prompt and response bodies are not written to this ledger by default.

Your first governed request is complete when its owner, route, budget, and evidence are understandable. Review the result, then add fallback, guardrails, and broader traffic one boundary at a time. The next guides cover each workflow in detail.

Put it into practice

See the working controls behind the guide.

Run a bounded provider simulation without a login, card, key, or model charge. Create a workspace only when you are ready to save controls or connect live traffic.

Create a free workspace Try the Test LabTalk with us