This is how you Optimise Microsoft Fabric Data Agents

Microsoft Fabric Data Agents are brilliant for data discovery. You can talk to your data without the need to know the questions you're going to ask upfront. But, out of the box Fabric Data Agents might not perform as great as you'd want. After all, it is based on a GPT-5 model and we all know that can misinterpret your prompt, or even hallucinate. In this video I'm going to show you how to Optimise your Fabric Data Agents. We will look at creating a multi data source agent that flawlessly picks the correct data source, understands the data model, and can answer complex questions with ease. We will do that by using the Agent Instructions, Data Source Descriptions, Data Source Instructions, and Sample Queries functionality. Sit back and learn how to make your Microsoft Fabric Data Agents behave. If you stay long enough, you might just find yourself learning how to have Fabric create you an apple pie recipe, and how to stop that.