Note
Access to this page requires authorization. You can try signing in or .
Access to this page requires authorization. You can try .
Connect Data Agents to your Azure Search Index in Microsoft Foundry (preview)
Data Agent creators can now connect their agents directly to Azure AI Search indexes built in Microsoft Foundry, unlocking powerful unstructured data scenarios. Using the resource URL, you can securely connect to your index—Data Agents fully respect the permissions of your Azure AI resources. In Foundry, you can craft rich AI Search indexes with custom enrichments, preprocessing logic, and tailored schemas for PDFs, text files, and more. Once connected, Data Agents can reason over that unstructured content and even join insights from your index with your structured data sources, giving you a unified, intelligent view across all your data.
Important
This feature is in preview.
Prerequisites
- A paid F2 or higher Fabric capacity, or a Power BI Premium per capacity (P1 or higher) capacity with Microsoft Fabric enabled.
- Enable cross-geo processing and cross-geo storing for AI based on requirements explained in Fabric data agent tenant settings.
- At least one of these data sources, with data: A warehouse, a lakehouse, a Power BI semantic model, a KQL database, a mirrored database, or an ontology. You must have read access to the data source.
Set up your Azure AI Search resource
To connect your Data Agent to an Azure AI Search index, first ensure your search resource is properly configured.
Create an Azure AI Search index.
Use the Azure AI Search quickstart to create an index. You can start with sample data or use your own.Enable role-based access control.
Turn on role-based authentication for your search service and index. This permission allows the Data Agent to securely access your resource using the identity of the asking user.Assign the required roles.
Confirm that your user or service principal has the following roles on the Azure AI Search resource:Search Index Data ContributorSearch Index Data Reader
These permissions ensure the Data Agent can read your index and retrieve relevant content.
Retrieve the resource URL.
Copy the resource URL for your Azure AI Search service. You need this value when adding the connection in your Data Agent configuration.
Tip
The Data Agent can include citations so you can see which documents were used to generate a response. Citations appear in the user experience only if at least one of the following fields is present (case-sensitive): url, sourceUrl, filePath, path, or folderPath.
To ensure citations display correctly in the Data Agent, you may want to include one of these fields in your index schema.
Connect index to data agent
Go to the Data tab and select Add AI Search Index.
Provide your resource URL when prompted.
Ask a question such as "Tell me more about the Uptown Chic hotel" to query your index.
View the documents used to generate the answer through the reasoning steps.
Note
When a user asks a question, the Data Agent sends the user's identity to the Azure AI Search index. This permission ensures that access controls and permissions defined on the index are respected.
Configure the index in your Data Agent
You can configure how your Data Agent uses your AI Search index. You can provide context about the source, give instructions on which fields to reference, and adjust parameters that control how much context is returned.
Context
Use this field to describe your AI Search index. Include details about what the index contains, key fields, and how it should be used. This helps the agent correctly route questions to the appropriate index.
👁 Screenshot of the Data Agent context.
Configuration
Use these settings to control how your Data Agent queries and interprets results from your AI Search index:
| Setting | Description |
|---|---|
| Display Name | The name displayed for this index within the Data Agent. |
| Search Type | Choose from the available search options supported by your index (for example: full-text, hybrid, or semantic search). |
| Number of Documents | Select how many documents the agent should retrieve per query. The recommended range is 3–20. Higher values may return more context but can increase processing time. |
👁 Screenshot of the Data Agent configuration.
Agent Instructions
Use the agent instructions to guide how the agent processes and composes the final answer. The search step retrieves the relevant document chunks, but the instructions tell the agent how to interpret that information, review it, and structure the response.
👁 Screenshot of the Data Agent instructions.
Next steps
Feedback
Was this page helpful?