What do I want to learn at FabCon this year? I have high expectations for the FabCon conference during the week of March 16-20, 2026. With over 30,000 organizations actively using Fabric in production, I expect to hear much more about what we, as a community, have learned about planning projects and implementing, monitoring and managing solutions. Of course, there will be marketing rhetoric about new and upcoming features. There will be announcements about preview features, and I’m interested in the long-term product vision, but I am anxious to learn more from others who are in the trenches and using Fabric day-to-day. I’m also intrigued to see Microsoft getting back into the SQL database conference business. The PASS organization was once the center of the Microsoft data platform community, but a lot of things changed after the pandemic. Time will tell what will happen on the SQL Server front, but will Fabric be the catalyst to bring the entire data community back together?
Visit Me in the Community Lounge
I will be in the community lounge expert’s area from 9 to 11 on Thursday answering questions about Fabric certifications. I have a new O’Reilly book on the PL-300 exam and recently delivered Fabric certification prep sessions for the DP-600 and DP-700 exams.
The Microsoft Fabric platform has been generally available for over two years but well over five years in active development and testing. Many of the foundational components and services that comprise Fabric have grown and matured for decades. As Fabric continues to expand and mature, what are we to expect to see at the Fabric and SQL Conference (FabCon & SQLCon) in March and in 2026?
Copilot, Power BI and “Chat with Your Data”
Many of our consulting clients want their users to use chat, much the same way they use external AI agents, to gain insights from secure organizational data. A collection of new Power BI features was recently added to enhance semantic models and Fabric data agents for optimized AI enablement, including verified, answers and AI instructions. Early testing last year was rough but recent experience shows significant improvement in chat with your data (CWYD), using agents and grounded semantic models.
- In practice, how is CWYD enabled across enterprises, tuned and optimized?
- What are we seeing as we have implemented Copilot and agentic AI chat experiences for our customers?
- How to we prepare to rollout Power BI CoPilot & CWYD at-scale?
- What are the expected and realized costs related to OpenAI token and Fabric CU usage as org users adopt these features?
Ontologies and Fabric IQ: The Next Frontier
Ontologies, a new workload in Fabric, promises to enable next-level semantic understanding of an organization’s business. They include but go beyond the holistic domain-level extension of semantic models and data agents and include operations agents that monitor the business in real time. Fabric IQ uses the Fabric Graph engine to define and apply conceptual business language to agent-based queries.
- What is the practical application for Fabric IQ for business users in different roles?
- How can we plan to layer operations agents and ontologies over data agents and prepared semantic models?
- How are ontologies planned, managed and governed?
- What are the performance and cost impacts for an organization?
SQL Database in Fabric: When and why
Fabric was introduced to the industry as an analytics platform. Fabric data warehouses, lakehouses and Power BI are all optimized for fast read-only performance at scale. Now, SQL Database for Fabric enters the picture with compatibility and most capabilities of on premises SQL Server.
- How does SQL Database in Fabric scale?
- What is the cost? How do we estimate cost within the Fabric CU billing model?
- When does it make sense to use an OLTP database in Fabric?
- Under what conditions does it make more sense to use Azure SQL, SQL Server on prem or SQL Server in a virtual machine?
RTI and Eventhouse: The Proliferation of realtime streaming data analysis
I’m very familiar with the real time intelligence features and capabilities.
- What are production customers experiencing in terms of capacity utilization and cost?
Grown-up ETL: Next level data transformation, orchestrations and streamlined data ingestion automation
Pipelines, notebooks and now Copy Job are the core workloads used in most enterprise scale Fabric data engineering projects. Among the data engineering community, Dataflows is often viewed as the analyst user’s self-service transformation tool, but this recent Fabric team blog post suggests that integration is moving forward.
- How can we manage object ownership and connection portability in large-scale, multi-environment solutions?
- As the product teams continue to introduce more convenient transformation features, how necessary are heavily orchestrated solutions?
- Is Dataflows Gen2 going to be an enterprise-ready component of Fabric Data Factory solutions or is it the self-service ETL for analyst Power BI users?
