The good, the bad, the ugly with Power BI agentic development

Power BI AI and agentic development are arguably the biggest trends in the Microsoft universe right now. Every day, our feeds are flooded with blogs and LinkedIn posts proclaiming the "magic" of LLMs and AI agents. Naturally, this has sparked a wave of concern: Are Power BI developers losing their purpose? Will we be replaced by these models?
The short answer: Not if you know how to use them.
Join me in this encouraging, deep-dive session where we explore how to profit from agentic development and identify the tools that will make you faster, not obsolete.
The Good
We’ll start with the "Good"—the immediate wins. Through live demos, we will explore how LLMs and MCP (Model Context Protocol) Servers can optimize your daily workflow. You’ll get a clear overview of how agentic development can handle the heavy lifting of report creation and data prep.
The Bad
However, it isn't all sunshine and automation. We will pull back the curtain on the "Bad"—the hard limits of AI. Where should you be careful? What are the inherent risks? We will tackle the critical question: Can we truly rely 100% on these tools for enterprise-grade analytics?
The Ugly
Why is "The Ugly" in the title? Because there is a growing risk of over-reliance. If we let AI do all the thinking, how do we ensure the next generation of developers actually learns the core concepts of for example DAX? Furthermore, we will address the "ugly" reality of data protection.
By the end of this Level 300 session, you will walk away with a balanced, realistic understanding of how to strategically implement AI in your daily work and across your organization.

Master agentic tools like MCP Servers to accelerate Power BI development. Identify critical AI limitations and data privacy risks to avoid "ugly" errors. Build a balanced strategy to integrate LLMs without losing core expertise.