Microsoft Fabric at Build 2026

Microsoft Build 2026 in San Francisco brought a wave of announcements around agentic development, and Microsoft Fabric sat right at the center of them. Whether you are a data engineer, an analytics developer, or someone building AI apps on top of your data estate, there is a session here for you.

Below is the shortlist I have been pointing folks to, organized by what you are trying to do. Most of these are available on-demand so you can watch on your own schedule. A few sessions (labs, table talks, and lightning talks) were not recorded, so I have flagged those with (Not recorded) below. If you missed them live, the session pages still have abstracts, speaker info, and in some cases follow-up resources.

Start here: agentic analytics

If you only have time for one session, make it this one. Emily Lisa and Sujata Narayana walk through how AI agents are reshaping the way we build semantic models, reports, and end-to-end analytics solutions on Power BI and Fabric.

• OD817 - Agentic analytics with Power BI and Microsoft Fabric (Pre-recorded, Level 200)

Building agentic apps on Fabric

This is where Rayfin, the new managed backend on Fabric, takes center stage. The breakout sets the vision, the demos show it in action, and the labs give you hands-on time.

• BRK225 - Data, apps, and agents: the future of app dev with Rayfin (Breakout, Level 300)

• DEM313 - Build agentic apps in minutes with Rayfin and Microsoft Fabric (Demo, Level 300)

• DEM362 - Building a Multi-Agent Workflow in Microsoft Fabric (Demo, Level 300)

• OD810 - Build fast, not fragile with Rayfin and Microsoft Fabric (Pre-recorded, Level 200)

• LAB514 - Ship AI apps fast with Rayfin, a new managed backend as a service (Lab, Level 300) (Not recorded)

• LAB513 - Build an AI app with Azure SQL Hyperscale, Microsoft Fabric, and Foundry (Lab, Level 300) (Not recorded)

• LTG458 - Agentic Apps on Microsoft Fabric: Mind-Blowing Interactive Experiences (Lightning Talk, Level 200) (Not recorded)

• TT677 - Agentic build experiences with Microsoft Fabric (Table Talk, Level 300) (Not recorded)

Analytics, BI, and the data warehouse

For analysts and data engineers who want to see where the platform is heading, these sessions cover the data warehouse, the AI-native data engineer experience, and applied data science on Fabric.

• OD813 - Powering modern data analytics in Fabric Data Warehouse (Pre-recorded, Level 200)

• OD818 - The AI-native Data Engineer (Pre-recorded, Level 200)

• DEM368 - Data Science and Machine Learning with Microsoft Fabric (Demo, Level 100)

Platform, data estate, and governance

If your job involves keeping the data estate unified, secure, and ready for AI, start here. Fabric IQ and Real-Time Intelligence are the standout topics this year.

• OD815 - Unify your entire data estate on a single, AI-ready data lake (Pre-recorded, Level 200)

• OD816 - Securing, scaling, and sustaining your data estate in Microsoft Fabric (Pre-recorded, Level 200)

• OD811 - Powering the next AI frontier with a unified data platform (Pre-recorded, Level 200)

• OD812 - Fabric IQ: Bringing enterprise intelligence into the developer workflow (Pre-recorded, Level 200)

• OD819 - Real-Time Intelligence: Building event-driven AI apps and agents (Pre-recorded, Level 200)

Fabric meets Foundry

Two lightning talks that connect the dots between Fabric data and Foundry agents. Short, punchy, and worth the 10 minutes each.

• LTG448 - Build Intelligent Agents with Work IQ, Foundry IQ, and Fabric IQ (Lightning Talk, Level 300) (Not recorded)

• LTG407 - From Fragmented Data to Agentic Intelligence at Enterprise Scale, Powered by Fabric and Foundry (Lightning Talk, Level 200) (Not recorded)

A suggested viewing order

If you want a learning path rather than a buffet (recorded sessions only, so you can binge on-demand):

1. Start with OD817 for the big picture on agentic analytics.

2. Watch OD815 and OD811 to ground yourself in the unified data platform story.

3. Move to BRK225 for the agentic app dev vision, then DEM313 and DEM362 for the demos.

4. Go deeper with OD818 (The AI-native Data Engineer) and DEM368 (Data Science and ML with Fabric) to see Fabric applied end to end.

5. Wrap up with OD812 (Fabric IQ) and OD819 (Real-Time Intelligence) for a look at where the platform is heading.

Note: the labs (LAB513, LAB514), table talk (TT677), and lightning talks (LTG458, LTG448, LTG407) were not recorded. If those topics interest you, the session pages still have abstracts and speakers, and the lab content is often republished as hands-on tutorials on Microsoft Learn.

Happy building. If you watch any of these and want to compare notes, reach out.