Using only agentic prompts, we’ll build a complete Power BI solution end-to-end—source queries, transformations, semantic model, relationships, measures, and report visuals. The whole kit and kaboodle.

Impressive? Sure. But where does agentic BI development actually deliver value—and where does it break down, requiring an experienced practitioner to step in?

In this session, we’ll use GitHub Copilot, modern LLMs, and the Power BI MCP Server to accelerate development in a realistic project scenario. Along the way, we’ll separate signal from noise, highlight where AI excels, and expose the gaps. More importantly, you’ll learn the skills required to guide these tools—prompting, validation, and design discipline—to produce solutions that are not just fast, but correct, scalable, and production-ready.