Skip to main content
cancel
Turn on suggestions
Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.
Showing results for 
Search instead for 
Did you mean: 

Did you hear? There's a new SQL AI Developer certification (DP-800). Start preparing now and be one of the first to get certified. Register now

chafia
‎03-24-2026 11:30 AM

What's next for Fabric IQ Ontology: The operational context that powers your AI agents (Preview)

If you haven’t already, check out Arun Ulag’s hero blog “FabCon and SQLCon 2026: Unifying databases and Fabric on a single, complete platform” for a complete look at all of our FabCon and SQLCon announcements across both Fabric and our database offerings. 

Since introducing Fabric IQ, we've seen growing momentum around how organizations are using ontologies to establish a shared semantic foundation for analytics and AI. By modeling business concepts and relationships in the ontology, teams can move beyond fragmented data views and enable AI experiences that reason, automate, and act with confidence.

At FabCon Atlanta, we shared a preview of what’s coming next: new Fabric IQ capabilities that make ontologies more expressive, more connected, and more reliable as the foundation for AI-powered experiences across Microsoft Fabric.

Why Fabric IQ and Ontology matter

AI is only as good as its understanding of your business. While Microsoft Fabric and OneLake unify where data lives, the meaning of that data often remains scattered across systems, tools, and teams. This leads to inconsistent definitions, brittle integrations, and AI outputs that are difficult to trust.

Fabric IQ changes that—at the center of Fabric IQ is the Ontology item, which connects data, processes, rules, and actions into a unified semantic layer. By binding real-world data to business entities and relationships, ontologies elevate raw tables and events into business-ready concepts that both people and AI can understand and use consistently.

What's coming next:

Highlights for FabCon Atlanta - Embedding rules and actions into our Ontology

Rules transform live business context in the ontology into operationalized outcomes. The ontology becomes capable of automatically initiating business processes through alerts and automated actions, all within a seamless and context-aware environment. This feature helps you unlock automation capabilities without needing to switch between tools or write custom code, to streamline workflows and enhance efficiency.

Refer to the Rules in ontology (Preview) (with Fabric Activator) documentation to learn more.

What_s_next_for_Fabric_IQ_Ontology_The_operational_context_that_powers_your_AI_aWhat_s_next_for_Fabric_IQ_Ontology_The_operational_context_that_powers_your_AI_a

Figure: Create a rule to get automatically notified when inventory drops below your threshold.

Enterprise-grade security and governance

As ontologies become mission-critical, enterprise controls are essential.

Enhanced access control and permissions management
Fabric IQ now supports sharing and permissions management for Ontology items, enabling you to control who can access and collaborate on ontologies. Using Fabric's permissions management experience, you can grant granular permissions such as read, edit, and reshare, and see who has access to Ontology items within a workspace. This ensures ontologies can be securely shared and governed as they become foundational to analytics and AI workflows across your organization.

Private network access with Azure Private Link
Fabric IQ now supports Azure Private Link integration, enabling organizations to set network isolation at the tenant and workspace levels. All connectivity to ontology items flows through secured endpoints, ensuring traffic never traverses the public internet. Cross-workspace validation ensures that isolated workspace artifacts can safely invoke artifacts from non-isolated workspaces, providing both security and flexibility for complex enterprise topologies.

Powering AI Agents with business context

At Ignite, we introduced data agents that leverage Ontology for consistent business definitions and relationships. We're now expanding this approach, allowing more agents to use Ontology as business context for enhanced scenarios beyond Q&A.

Refer to the Ontology (preview) tutorial part 4: Create data agent documentation to learn more.

Ontology integration with Operations Agent (Coming Soon)

Operations Agent will be able to use your Ontology as a knowledge source to inform its playbooks. By understanding entity types, relationships, rules, and actions defined in the ontology, agents can evaluate conditions and take actions that align with your business logic, grounded in shared semantics rather than isolated signals.

Refer to the What is ontology (preview)? documentation to learn more.

Ontology MCP endpoints (Coming Soon)

Fabric IQ Ontology will be accessible to a broader agent ecosystem through public Model Context Protocol (MCP) endpoints. Once available, these endpoints will enable agents across Microsoft’s ecosystem and beyond to ground reasoning and actions in a unified semantic layer reinforcing the ontology as a central source of business understanding for AI.

Refer to the What is ontology (preview)? documentation to learn more.

Learn more, engage, and stay updated

Fabric IQ and Ontology continue to evolve, and FabCon Atlanta marks another step forward in making first-class business semantics across analytics and AI in Microsoft Fabric.

To stay engaged:

0