This course introduces using AI to develop Power BI semantic models and reports. This short video course has nine modules and shows what you can di with an AI agent in a Power BI model, what you need to provide to get correct results, and what knowledge you need to review the results.
We start by showing that a business question typed into a popular web AI chat requires many additional details to generate a valid DAX measure. In contrast, the same question provides an immediate response using AI tools connected to the semantic model itself. The methodology is the same across vendors, even though the tools and their setup instructions could change change.
The core of the course is giving the agent access to the semantic model. An MCP server lets the agent inspect tables, columns, relationships, and measures, and create objects in the model after your approval. We show the Power BI modeling servers and create a measure on the Contoso model. We document choices about the calendar, currency conversion, customer counting, and DAX standards in files the agent can read. By using skills, you reuse this guidance in other semantic models. Finally, we introduce the report authoring skill, which can modify the report definition once you save the file as a PBIP project.
The companion content available to the students includes:
PBIX files for all the demos
Prompts using in the demos
Students have access to a private discussion area where they can interact with the instructors, asking questions related to the lectures and the exercises.