Fabric IQ

Fabric IQ – Introduction, Architecture & Full Hands-On Workshop In this episode of Fabric Friday, we explore Fabric IQ, the new ontology-based semantic foundation announced at Microsoft Ignite. Fabric IQ in the Fabric Periodic Table: https://www.fabricperiodictable.com/#element-fabric-iq Fabric IQ extends traditional semantic models by defining entities, properties and verb-based relationships (e.g., Store has SalesEvent). This structure gives LLMs deeper understanding and enables powerful reasoning across domains. What’s inside the video: 00:00 - Introduction to Fabric IQ and agenda overview 00:01 - OneLake foundation and data sources for Fabric IQ 00:02 - Concept of ontology and enterprise vocabulary 00:04 - Different ways to use your ontology (visualize, query, operations agent) 00:07 - Defining entities, metadata, and relationships in ontology 00:11 - Example business model with store, freezer, and temperature data 00:17 - When to use ontology vs semantic models decision tree 00:22 - End-to-end workflow: design, bind, generate, query, integrate 00:31 - Practical lab: creating Lakehouse and Eventhouse data sources 00:47 - Running queries with data agents and demonstration wrap-up Key Takeaways: • Ontologies create enterprise-wide vocabularies • Ideal for cross-domain reasoning, complex relationships and real-time scenarios • Works best with Lakehouse + Eventhouse + Direct Lake • Ontology & semantic models complement each other • Preview features require iterative testing and validation If you enjoy this deep dive, subscribe for new Fabric Friday episodes every week — always focused on architecture, governance and real-world Fabric implementations.