Most Power BI estates are full of columns and measures nobody uses and nobody dares to delete. In this session, you'll see how to catalogue the tables, columns and measures across your semantic models, map which ones reports actually use, and turn the results into a safe cleanup plan built around your own rules, not a generic checklist.

You'll also see where AI earns its place (triaging findings, drafting the plan, explaining the results to stakeholders) and where it shouldn't be trusted, with a simple rule to take back to your team: let deterministic checks find it, let AI help you decide and explain it, and let a human approve what changes.