Explicit Measures Podcast 508: Kicking Off Fabric the Right Way
Mar 5, 2026
They tackle how to start Fabric with a clean semantic model and avoid fragmented, redundant models. They debate rebuild vs incremental consolidation and how to use AI to audit and merge existing models. They cover governance, workspace and lakehouse design, access controls, training, and strategies to shape team culture toward a single source of truth.
AI Snips
Chapters
Transcript
Episode notes
Migrate Incrementally While Keeping Production Running
- Migrate incrementally; don't freeze production work but reallocate resources to build a gold semantic model from top used reports.
- Tommy suggests prioritizing a smaller production footprint and dedicating people to consolidation tasks.
AI Makes Model Consolidation Fast And Low Risk
- AI lowers the barrier to consolidate multiple models by quickly surfacing redundancies and candidate central tables.
- Mike describes using agents to merge PBIX/PBIP files and get an initial prioritized list of duplicate tables and measures.
Provide Approved AI Tools To Avoid Shadow IT
- Sanction and provide approved AI tools to your team to prevent shadow IT and data leakage.
- Mike warns that if leadership blocks AI, employees will use unsanctioned tools and risk sensitive data exposure.