MyFabric Workshop: From Data to Agents – AI in Microsoft Fabric (EN)

How Does Fabric Knowledge Become the Enterprise Context for Copilot and Business Agents?

Online workshop | December 10, 2026 | 9:00 AM CET

AI agents are only as good as the context they have. In this workshop, we show how Microsoft Fabric uses Fabric IQ to turn enterprise data into that context, built on OneLake, semantic models and ontologies.

After a concise introduction to business agents and the Fabric platform, we get hands-on: you build a data agent that answers business questions in natural language and an operations agent that monitors real-time data and triggers actions. To wrap up, we look ahead at integration with Microsoft 365 Copilot and Copilot Studio and what this means for your data strategy.

Learning Objectives

• Understand what defines business agents and why data context determines their quality
• Know the building blocks of Fabric IQ (OneLake, semantic model, ontology, graph, data agent and operations agent) and how they work together
• Set up a data agent on a lakehouse, semantic model or ontology, then refine it with instructions and test it
• Configure an operations agent for real-time data that evaluates rules and recommends or triggers actions
• Realistically assess prerequisites, governance and costs (capacity, tenant settings, preview status)
• Identify initial use cases for your own organization

Prerequisites

To take an active part in the hands-on sessions, you need a laptop with an up-to-date browser and access to a Fabric workspace on a paid Fabric capacity (F2 or higher; a trial capacity is not sufficient). The following tenant settings must be enabled: Copilot/Azure OpenAI, Data Agent, Operations Agent and Ontology (Preview). You also need a Microsoft Teams account. Basic knowledge of Power BI or Fabric is helpful.

You can also join passively: all hands-on steps are shown live on screen.