Treza LabsT R E Z A

Connect Azure OpenAI to Treza

Connecting Azure OpenAI takes a little more than OpenAI: besides an API key, Azure routes requests to a specific resource and deployment, so Treza needs your full upstream URL too. Treza stores your key encrypted and uses it server-side to forward redacted requests — it's never exposed to your agents.

Create an Azure OpenAI resource and deployment

  1. In the Azure portal, create an Azure OpenAI resource in a supported region. (Access may require an application, depending on your subscription.)
  2. Open the resource in Azure AI Foundry (formerly Azure OpenAI Studio) and deploy a model — for example gpt-4o. Note the deployment name you choose; you'll need it in the URL.

For the full walkthrough, see Microsoft's Azure OpenAI quickstart.

Find your key and endpoint

  1. In the Azure portal, open your resource and go to Keys and Endpoint.
  2. Copy KEY 1 — this is your Azure OpenAI API key.
  3. Copy the Endpoint — it looks like https://your-resource.openai.azure.com.

Add it to Treza

  1. In the Treza dashboard, open Redaction proxies → New proxy and set Upstream provider to Azure OpenAI.
  2. Paste your key into the Azure OpenAI API key field.
  3. In Upstream URL, enter the full chat completions URL for your deployment:
https://your-resource.openai.azure.com/openai/deployments/your-deployment/chat/completions?api-version=2024-10-21

Replace your-resource, your-deployment, and the api-version with your own values.

  1. Choose which PII types to redact, then create the proxy.

Your provider key is encrypted at rest and only used to forward requests upstream. Treza never returns it to your agents or includes it in logs.