--- title: "OpenAI setup in Turbo360" slug: "openai-setup-in-turbo360" updated: 2026-10-29T11:13:41Z published: 2026-10-29T11:13:41Z canonical: "docs.turbo360.com/openai-setup-in-turbo360" --- > ## Documentation Index > Fetch the complete documentation index at: https://docs.turbo360.com/llms.txt > Use this file to discover all available pages before exploring further. # OpenAI setup in Turbo360 ## Overview Turbo360 supports AI-powered features through Azure OpenAI integration. This integration enables AI-generated summaries and resource analysis within Cost Analyzer and Business Applications, and AI-generated explanations within Tenant Manager. The setup experience differs based on your deployment model: - **SaaS:** AI capabilities are available by default and do not require additional configuration. - **Private hosting:** AI capabilities require a customer-managed Azure OpenAI resource to be configured. This article covers the full setup for private hosting deployments. ### Business value Connecting Turbo360 to your own Azure OpenAI resource enables AI-generated cost summaries, resource analysis, and AI Agent responses within Cost Analyzer, Business Applications, and Tenant Manager, without routing data through a shared service. This gives private hosting customers full control over their AI model, region, and data residency. ## Prerequisites - An active Azure subscription with access to Azure OpenAI - Permission to create resources and assign roles in the target resource group - A completed Turbo360 private hosting deployment ### Required permissions | Action | Required role | | --- | --- | | Create Azure OpenAI resource | Contributor on the target resource group | | Deploy a model in Azure AI Foundry | Contributor on the Azure OpenAI resource | | Configure OpenAI connection in Turbo360 | Turbo360 organization admin | ### Dependencies - Azure OpenAI must be available in your target Azure region. Not all regions support all model types. - The deployed model must use a compatible deployment type: **Global Standard**, **Standard**, or **DataZone Standard**. ## How it works Turbo360 connects to Azure OpenAI using an API key and endpoint URL. Once the connection is saved, Turbo360 uses the configured model for all AI Agent requests within Cost Analyzer, Business Applications, and Tenant Manager. In Tenant Manager, these requests come from the **AI agents** tab of a user or license record and from the **Explain** button in recommendation details. On private hosting, they need this connection to be saved. The connection is scoped to the organization and applies to all users. SaaS deployments use a Turbo360-managed OpenAI connection and do not require any configuration. Private hosting deployments must provide their own Azure OpenAI resource. ### AI agents account-wide toggle The Account owner can enable or disable AI agents for the entire organization from a single toggle. When disabled, AI agents are hidden across the app for every user, regardless of their individual role permissions. The toggle location depends on your deployment model: - **SaaS:** navigate to **Settings** > **General settings**. The **AI agents** toggle is in the right panel. - **Private hosting:** navigate to **Settings** > **OpenAI connection**. The **AI agents** toggle appears on the same page as your connection configuration. This toggle controls visibility at the account level. It does not override or replace the per-role AI Agents permission. Both must be enabled for a user to access AI agents. ## Steps Use the following steps to create an Azure OpenAI resource, deploy a compatible model, and connect it to Turbo360. Navigate to the Azure portal to get started. ### Create the Azure OpenAI resource Creating an Azure OpenAI resource provisions the service endpoint your Turbo360 instance will connect to. 1. Sign in to the Azure portal. 2. Search for **Azure OpenAI** and select **Create**. 3. Provide the required details: **Subscription**, **Resource Group**, and **Region**. 4. Complete the resource deployment. ### Deploy a model Deploying a model creates the specific AI model instance that Turbo360 will call when generating responses. Note: Compatible deployment types for Turbo360: **Global Standard**, **Standard**, and **DataZone Standard**. 1. Open the Azure OpenAI resource. 2. Go to **Overview** > **Explore Foundry portal**. 3. In the Microsoft Azure AI Foundry portal, go to **Shared resources** > **Deployments**. 4. Click **+ Deploy model** > **Deploy base model**. 5. Select the model and click **Confirm** to start deployment. ### Retrieve connection details Retrieving the connection details gives you the values required to configure the OpenAI connection in Turbo360. 1. From the **Model deployments** screen in the Azure AI Foundry portal, click the deployed model. 2. Note the following values. You will need them in the next step: - **API key** - **Endpoint URL** - **Deployment name** ### Configure the OpenAI connection in Turbo360 Configuring the connection in Turbo360 activates Azure OpenAI for all AI Agent features in your private hosting instance. 1. In Turbo360, navigate to **Settings** > **OpenAI connection**. 2. Click **Create**. 3. Enter the connection details retrieved in the previous step: - **Endpoint URL** - **API key** - **Deployment name** 4. Click **Validate** to verify the connection. 5. Confirm the validation is successful, then click **Save**. ## Example scenario A platform engineer at an enterprise running Turbo360 as a private hosted instance creates an Azure OpenAI resource in West Europe, deploys a GPT-4o model using the Global Standard deployment type, and retrieves the API key and endpoint URL from the Azure AI Foundry portal. They navigate to **Settings** > **OpenAI connection** in Turbo360, enter the connection details, validate successfully, and save. Cost Analyzer users in the organization can now use AI Agents to generate cost summaries and resource analysis responses. ## Limitations - VNet-enabled private hosting instances must ensure the Turbo360 Function Apps can reach the Azure OpenAI endpoint. If the OpenAI resource is behind a private endpoint, outbound VNET integration must be configured to allow connectivity. - Only one Azure OpenAI connection can be configured per Turbo360 organization. Changing the connection applies to all users and modules immediately. - SaaS deployments cannot configure a custom Azure OpenAI connection. The Turbo360-managed connection is used automatically. - API keys are stored at the organization level. Rotate keys in Azure and update the connection in Turbo360 to avoid AI feature interruptions. ## Troubleshooting 1. **Invalid API key error on validation** Cause: The API key entered does not match the key for the configured Azure OpenAI resource. Fix: Open the Azure OpenAI resource in the Azure portal, go to **Keys and Endpoint**, and copy the correct key. Update the connection in Turbo360 and validate again. 2. **Invalid endpoint error on validation** Cause: The endpoint URL is malformed or points to the wrong Azure OpenAI resource. Fix: Confirm the endpoint URL from the Azure AI Foundry portal under the deployed model's details. The URL should follow the format `https://<resource-name>.openai.azure.com/`. 3. **Deployment not found error on validation** Cause: The deployment name entered in Turbo360 does not match the name of the deployed model in Azure AI Foundry. Fix: Open the Azure AI Foundry portal, go to **Shared resources** > **Deployments**, and confirm the exact deployment name. Names are case-sensitive. 4. **Validation failed: connection cannot be established** Cause: Network connectivity between Turbo360 and the Azure OpenAI endpoint is blocked, or the Azure OpenAI resource's firewall restricts access. Fix: Check the Azure OpenAI resource's networking settings. If the resource uses a private endpoint or IP restrictions, ensure the Turbo360 Function Apps are allowed through the firewall or are connected via VNET integration. 5. **AI Agents tab does not appear after connection is saved** Cause: The connection was saved but the module cache has not refreshed. Fix: Reload the Cost Analyzer, Business Applications, or Tenant Manager page. If the tab is still absent, navigate to **Settings** > **OpenAI connection** and confirm the connection status shows as valid. 6. **AI agents are not visible to any user despite role permissions being correct** Cause: The account-wide AI agents toggle has been disabled by the Account owner. Fix: Navigate to **Settings** > **OpenAI connection** (Private hosting) or **Settings** > **General settings** (SaaS) and confirm the **AI agents** toggle is enabled. ## Related articles - [AI Agents in Business Applications](https://docs.turbo360.com/docs/ai-agents-in-business-applications) - [AI Agents in Cost Analyzer](https://docs.turbo360.com/docs/ai-agents-cost-analyzer) - [Explore users](https://docs.turbo360.com/docs/explore-users-in-tenant-manager) - [Explore licenses](https://docs.turbo360.com/docs/explore-licenses-in-tenant-manager) - [Optimization overview](https://docs.turbo360.com/docs/optimization-overview-in-tenant-manager) - [Network configuration](https://docs.turbo360.com/docs/network-configuration) - [VNet configuration during deployment](https://docs.turbo360.com/docs/vnet-configuration-during-deployment) - [Private hosting architecture](https://docs.turbo360.com/docs/private-hosting-architecture)