updated module learn.wwl.implement-fabric-data-agents

Author: weslbo

learn-pr/wwl/implement-fabric-data-agents/azure-ai-foundry.yml

@@ -4,7 +4,7 @@ title: Integrate Microsoft Fabric data agents with Microsoft Foundry
 metadata:
   title: Integrate Microsoft Fabric Data Agents With Microsoft Foundry
   description: "Integrate Microsoft Fabric data agents with Microsoft Foundry"
-  ms.date: 08/18/2025
+  ms.date: 03/27/2026
   author: weslbo
   ms.author: wedebols
   ms.topic: unit

learn-pr/wwl/implement-fabric-data-agents/data-agent-capabilities.yml

@@ -4,7 +4,7 @@ title: Understand Microsoft Fabric data agent capabilities
 metadata:
   title: Understand Microsoft Fabric Data Agent Capabilities
   description: "Understand Microsoft Fabric data agent capabilities"
-  ms.date: 08/18/2025
+  ms.date: 03/27/2026
   author: weslbo
   ms.author: wedebols
   ms.topic: unit

learn-pr/wwl/implement-fabric-data-agents/end-to-end-scenario.yml

@@ -4,7 +4,7 @@ title: Implement an end-to-end Microsoft Fabric data agent
 metadata:
   title: Implement an End-to-End Microsoft Fabric Data Agent
   description: "Implement an end-to-end Microsoft Fabric data agent"
-  ms.date: 08/18/2025
+  ms.date: 03/27/2026
   author: weslbo
   ms.author: wedebols
   ms.topic: unit

learn-pr/wwl/implement-fabric-data-agents/exercise-copilot-fabric-data-agents.yml

@@ -4,7 +4,7 @@ title: Exercise - Chat with your data using Microsoft Fabric data agents
 metadata:
   title: Exercise - Chat With Your Data Using Microsoft Fabric Data Agents
   description: "Exercise - Chat with your data using Microsoft Fabric data agents"
-  ms.date: 08/18/2025
+  ms.date: 03/27/2026
   author: weslbo
   ms.author: wedebols
   ms.topic: unit

learn-pr/wwl/implement-fabric-data-agents/includes/azure-ai-foundry.md

@@ -13,8 +13,19 @@ The integration also reduces the steps required to **connect AI agents to enterp
 Before you begin, make sure that:
 
 - You have created and **published** a Fabric data agent endpoint.
-- Developers and end users in Microsoft Foundry have at least the Azure AI User **RBAC role** assigned.
-- Developers and end users have at least **read access** to both the Fabric data agent and its underlying data sources.
+- Developers and end users in Microsoft Foundry have at least the **Azure AI User** RBAC role assigned.
+- Developers and end users have at least **read access** to the Fabric data agent.
+- Developers and end users have the minimum permission on each underlying data source:
+
+  | Data source | Minimum permission |
+  |---|---|
+  | Power BI semantic model | **Build** (includes Read). Read alone isn't sufficient because the agent generates model queries that require Build. |
+  | Lakehouse | Read on the lakehouse item |
+  | Warehouse | Read (SELECT on relevant tables) |
+  | KQL database | Reader role on the database |
+
+> [!IMPORTANT]
+> This integration uses **identity passthrough** (On-Behalf-Of). The agent runs queries using the signed-in user's identity. Service principal authentication isn't supported for the Fabric data agent.
 
 ## How to integrate Microsoft Foundry with Fabric Data Agents?
 

learn-pr/wwl/implement-fabric-data-agents/includes/data-agent-capabilities.md

@@ -21,14 +21,16 @@ Fabric data agents can reason over multiple data sources, including:
 - Power BI semantic models.
 - Eventhouse KQL Databases.
 - Lakehouses and Warehouses.
+- Ontologies (for semantic, business-term-aligned queries).
+- Azure AI Search indexes (preview, via a connected Microsoft Foundry index, enabling reasoning over unstructured content).
 
 The Fabric data agent can only access data that you provide, with up to five data sources. For example, a configured Fabric data agent could include a mix of two Power BI semantic models, one lakehouse, and one KQL database. Within those data sources, you can select the relevant tables. 
 
 > [!IMPORTANT]
 > 
-> The Fabric data agent works best with 25 or fewer tables selected across all data sources. 
+> The Fabric data agent works best with 25 or fewer tables selected across all data sources.
 > 
-> You can't use the Fabric data agent to access unstructured data resources like .pdf, .docx, or .txt files.
+> The Fabric data agent doesn't directly access unstructured files (such as .pdf, .docx, or .txt) as native data sources. To reason over unstructured content, connect an Azure AI Search index built in Microsoft Foundry as an additional source (preview).
 
 ## Integration inside and outside of Fabric
 

learn-pr/wwl/implement-fabric-data-agents/includes/power-bi.md

@@ -8,6 +8,9 @@ Copilot in Power BI enables users to ask natural language questions and receive
 - Access to Copilot in Power BI
 - Required permissions for the relevant Fabric data agents and Power BI items
 
+> [!NOTE]
+> When using the Copilot search experience (where Copilot automatically scans and suggests data agents), the **Standalone Copilot experience** must be enabled in Power BI tenant settings (**Tenant settings** > **Copilot** > **Standalone Copilot experience**).
+
 ## Ways to Use Fabric data agents in Copilot
 
 There are two ways you can use Fabric data agents with Power BI. Either, you use Copilot search to find and invoke a Fabric data agent, or, if you already know which data agent to use, you can manually add that data agent to the Copilot session.
@@ -36,7 +39,7 @@ When you interact with a Fabric data agent through Copilot in Power BI, the foll
 
 2. **Send the query:** The question is sent to the selected Fabric data agent.
 
-3. **Answer retrieval:** The data agent identifies the most relevant data source (such as a lakehouse, warehouse, semantic model, or KQL database) and queries it. Data security protocols like Row-Level Security (RLS) and Column-Level Security (CLS) are enforced based on your permissions.
+3. **Answer retrieval:** The data agent identifies the most relevant data source (such as a lakehouse, warehouse, semantic model, KQL database, ontology, or Azure AI Search index) and queries it. Data security protocols like Row-Level Security (RLS) and Column-Level Security (CLS) are enforced based on your permissions.
 
 4. **Response delivery:** The Fabric data agent sends the answer back to Copilot.
 

learn-pr/wwl/implement-fabric-data-agents/includes/summary.md

@@ -1,6 +1,6 @@
 **Microsoft Fabric data agents** are AI-based tools that enable users to interact with enterprise data using natural language. They simplify access to information stored in Fabric OneLake by allowing users to ask plain English questions and receive responses based on the underlying data.
 
-With Microsoft Fabric data agents, users can ask questions about their data in plain English, without needing to know technical languages like SQL, DAX, or KQL. The platform provides a chat interface with debugging tools to refine and improve agent responses. Data agents can connect to a variety of data sources, including Power BI models, Eventhouse KQL databases, Lakehouses, and Warehouses, supporting analysis across multiple platforms. They also integrate with Fabric and external services such as Microsoft Teams, Copilot Studio, Microsoft Foundry, and custom applications.
+With Microsoft Fabric data agents, users can ask questions about their data in plain English, without needing to know technical languages like SQL, DAX, or KQL. The platform provides a chat interface with debugging tools to refine and improve agent responses. Data agents can connect to a variety of data sources, including Power BI semantic models, Eventhouse KQL databases, Lakehouses, Warehouses, and ontologies, supporting analysis across multiple platforms. With a connected Azure AI Search index (preview), agents can also reason over indexed unstructured content. They also integrate with Fabric and external services such as Microsoft Teams, Copilot Studio, Microsoft Foundry, and custom applications.
 
 ## Best practices
 

learn-pr/wwl/implement-fabric-data-agents/index.yml

@@ -3,7 +3,7 @@ uid: learn.wwl.implement-fabric-data-agents
 metadata:
   title: Implement Microsoft Fabric Data Agents
   description: Implement Microsoft Fabric Data Agents (chat with your data)
-  ms.date: 08/18/2025
+  ms.date: 03/27/2026
   author: weslbo
   ms.author: wedebols
   ms.topic: module

learn-pr/wwl/implement-fabric-data-agents/introduction.yml

@@ -4,7 +4,7 @@ title: Introduction
 metadata:
   title: Introduction
   description: "Introduction"
-  ms.date: 08/18/2025
+  ms.date: 03/27/2026
   author: weslbo
   ms.author: wedebols
   ms.topic: unit