If you haven't already, check out FabCon and SQLCon 2026 in Barcelona: Building the data foundation for Microsoft Copilot and Agents for a complete look at all of our FabCon Europe announcements across Fabric and Microsoft Databases.
Microsoft IQ: One enterprise, acting on one live picture
Organizations generate vast amounts of information every day across business systems, operational data, enterprise knowledge, communications, and the web. Yet the context needed to understand what is happening, what it means, and what should happen next often remains fragmented across teams, applications, and data sources.
Microsoft IQ brings that context together. It provides the intelligence layer that helps people and agents understand the world, the organization, and the business they serve by connecting knowledge, relationships, history, policies, and actions into a shared understanding.
Consider a global manufacturing operation days before a major product launch. Across factories, suppliers, warehouses, and transportation networks, thousands of decisions are being coordinated to meet customer commitments and keep the launch on track.
Suddenly, a critical production line goes down, putting delivery timelines, customer commitments, and business outcomes at risk. What happens next depends on far more than the failed equipment. The right response requires a complete understanding of the business, including maintenance history, production schedules, inventory, available capacity to supplier commitments, transportation options, launch milestones, policies, customer expectations, and changing external conditions.
No single team, application, or data source holds that picture. Instead, the enterprise must bring together all relevant signals, understand how they relate to one another, and coordinate the best response in real time. It must evaluate alternatives across factories, warehouses, suppliers, and transportation networks, balancing timing, cost, quality, risk, and customer commitments to determine the best path forward.
Microsoft IQ provides the intelligence layer for this type of challenge.
What AI changes and what intelligence it requires
AI is changing the economics, scale, and speed of decision-making. Choices that were once too small, too time-sensitive, too numerous, or too costly to evaluate can now be considered continuously, as agents monitor conditions, reason through alternatives, coordinate with people and other agents, and act within delegated authority.
At the same time, expectations are rising. Organizations are increasingly expected to anticipate disruption, adapt in real time, and meet commitments without exposing customers to the operational complexity behind the scenes.
That becomes difficult when data, business context, policies, and actions remain fragmented across systems. Agents can optimize individual functions, but they struggle to reason across the full business, often operating from different versions of reality. As a result, people are left stitching together context, reconciling conflicting answers, and coordinating decisions manually.
AI needs more than access to isolated applications or the latest rows of data. It needs one live, enterprise-wide model that combines current state, history, relationships, policies, governed decisions, live web data, and available actions. The defining question is simple: What does your AI run on?
Microsoft IQ expands the context available to people and agents by bringing together four complementary sources of intelligence:
- Web IQ grounds AI in current knowledge from the public web.
- Work IQ brings context from organizational knowledge and the flow of work.
- Foundry IQ connects AI to curated knowledge across structured and unstructured enterprise sources.
- Fabric IQ adds governed business meaning across data, analytical models, operational signals, entities, relationships, history, and available actions.
Together, the four IQs give AI a more complete understanding of the world, the organization, and the business it serves. Fabric IQ contributes the governed business context that connects this broader intelligence to the live state of the enterprise.
From Microsoft IQ to the Fabric operational nervous system
Within Microsoft IQ, Microsoft Fabric provides the live data foundation and operational loop that turns shared intelligence into action. Only Fabric brings real-time state, governed business context, and action together in one live picture, so people and agents can understand what is happening, decide what to do, and act from the same trusted foundation. As the unified platform for building an operational nervous system, Fabric enables the enterprise to sense, understand, reason, decide, act, and learn:
- Observe: OneLake unifies operational, analytical, and business data, while Real-Time Intelligence continuously detects emerging conditions and maintains a live picture of operations based on signals it’s connected to.
- Analyze: Fabric IQ adds business meaning through entities, relationships, history, rules, and available actions, helping transform signals into operational understanding.
- Decide: AI evaluates alternatives using business data and context, while governance policies, permissions, thresholds, and delegated authority remain embedded in every recommendation.
- Act: People and agents coordinate around the recommended response, with approved actions flowing into business systems and outcomes becoming part of the enterprise’s memory to inform future decisions.
The result is not another dashboard. It is a continuous operating loop that connects signals to understanding, decisions, and action. Fabric provides the operational foundation within Microsoft IQ, contributing the live business context, enterprise data, relationships, history, and actions that help people and agents reason over the current state of the business. Together, Fabric IQ, Web IQ, Work IQ, and Foundry IQ help create a more complete picture of the world, the organization, and the business they serve.
What Fabric makes possible
The value of an operational nervous system becomes clear when conditions change. A single operational signal can quickly ripple across manufacturing, supply chain, logistics, and customer commitments, requiring the enterprise to understand, decide, and respond as one.
When a critical production line goes down days before a product launch, Real-Time Intelligence detects the issue as it happens. Fabric immediately connect that signal to the broader context, including production plans, inventory, supplier commitments, transportation options, business communications, and external events. Fabric IQ then helps AI understand not only what happened, but what it may mean for the business.
With that context in place, AI can evaluate alternative responses, weighing timing, cost, quality, risk, and customer commitments. Leaders and agents are able to operate from the same shared understanding, with routine actions proceeding automatically where policy allows and higher-impact decisions are surfaced with explanations of clear tradeoffs and recommendations.
In this scenario, AI recommends reserving capacity at a second plant while production, materials, and transportation plans are rebalanced. Once approved, teams and agents execute against a coordinated response, with every recommendation, decision, and action remaining governed and traceable.
And the process does not end with a single decision. As conditions change because of supply disruptions, weather events, or shifts in demand, the same operational loop continues running, helping the enterprise sense change, reason across business context, govern decisions, and coordinate action.
Figure: When conditions change, Fabric helps organizations connect operational signals with business context, evaluate alternatives, and coordinate responses across the enterprise in real time.H&M is building its operational nervous system with Fabric
The shift is already beginning to take shape across industries.
For a global retailer like H&M, lost and stolen clothing is not simply an inventory challenge. It represents millions of euros in potential losses and underscores the need for greater real-time visibility into what is happening across the business.
H&M is connecting real-time signals across its operations to better track clothing and identify where items may be lost or stolen. This scenario demonstrates how bringing real-time data with business context can help teams move from fragmented signals to a clearer understanding of what is happening, why it matters, and where action may be needed.
This is a powerful example of how organizations can use real-time intelligence to address high-value business challenges and turn operational signals into measurable impact.
" Across our 4,000+ stores, some have ceiling readers that update every few minutes, others count with handheld scanners once a day, so the data was hard to compare. With Fabric Real-Time Intelligence we bring it all together into a fuller picture of where every garment is and what state it's in. We're using it for operational analytics and exploring how AI agents can make shopping easier." – Jan Barrish, Head of Retail Tech, H&M.
At FabCon, we are advancing the Fabric capabilities that help organizations connect signals, maintain operational understanding, and turn insight into governed action.
What's new in Fabric: Advancing the operational nervous system
The latest Fabric capabilities strengthen every stage of the operational nervous system, from connecting live signals and maintaining business context to enabling intelligence, analytics, and action.
Fabric Observability Insights (Preview)
Fabric Observability Insights enters public preview, using an AI-powered, domain-specific operations agent to investigate issues such as pipeline failures with Fabric-native context. The Fabric monitoring hub continues to provide a centralized view of job health, progress, history, and outcomes, while workspace monitoring adds log-level visibility in a secure, read-only Eventhouse database for analysis with Kusto Query Language (KQL) or SQL. To learn more, check out the workspace monitoring overview.
Connect and stream
Bring more operational data into Fabric with custom streaming connectors and enterprise-ready CDC support through the Mirrored Database Change Feed Connector. Schema-aware Eventstreams make it easier to process structured and evolving event data, while Reference Data Join, Copy job integration, and connector security extend real-time pipelines to richer operational scenarios.
Understand and analyze
Move from raw event data to useful answers faster with Copilot Data Exploration, which supports natural-language investigation in Real-Time Dashboards. Eventhouse MCP Server and Eventhouse Onboarding Agent extend that experience to AI-assisted access and guided solution creation, while shortcuts and update policies reduce duplication and maintenance.
Visualize and monitor
See operational conditions in context by bringing Fabric Maps into Real-Time Dashboards and connecting maps to external geospatial services through Feature Service Support. Copilot dashboard creation reduces the effort needed to create queries and visuals from a business question, while embedding and richer KPI and live-update experiences help bring operational insight into business workflows.
Act and automate
Turn important business moments into coordinated action with Business Events, giving applications, workflows, and agents a shared signal they can consume. Activator detects conditions and publishes or manages rule-driven responses, while Operations Agent extends the story into investigation and root-cause analysis.
What’s new in Fabric IQ
Investigate (Preview)
When the operations agent detects an anomaly in your business data and your Fabric observability data, it automatically runs Investigator to provide more context about what happened. Investigator analyzes telemetry data and detects correlated or explanatory patterns around the time of the anomaly, helping you understand the potential causes without manually investigating the data. The agent uses Fabric-native context to deliver evidence-backed hypotheses, impact assessment, and recommended next steps, simplifying and automating troubleshooting for data engineers. To learn more, check out the Operations Agent Actions document page.
Operations agent performance view (Preview)
Keep track of your operations agent’s behavior with observability information about the monitoring and reasoning steps it takes. Track the LLM usage as the agent spots issues in your business, wakes up, reasons over the data, and recommends next steps. To learn more, check out the Create and configure operations agents page.
Unified Semantic Understanding
Build a shared business model with Ontology AI generation and query, letting teams create ontology definitions and ask questions using business terminology. Semantic Models as a data source, Ontology Metrics, and mirrored database and SQL DB bindings connect governed measures and source data without duplicating definitions. Namespaces, relationships without joins, keyless entity bindings, complex relationship modeling, reuse and inheritance, and the full canvas experience make large, cross-domain models easier to organize and evolve. RDF import and export, versioning, resource links and enrichment, and the new Overview and Instances views help teams reuse existing knowledge, understand model changes, and work with ontology data at scale.
By working with Snowflake, we are helping customers reuse semantic context across platforms while advancing interoperability for semantic models and ontologies. By affirming our commitment to Apache Ossie (incubating), a community-led standard for exchanging semantic metadata, we are helping extend business meaning across analytics, AI, and BI ecosystems. This gives organizations a more open foundation for carrying consistent business meaning across the tools, platforms, and AI experiences they choose.
Figure: An ontology creates a shared understanding of the business by organizing data into meaningful entities, relationships, rules, and actions that people and AI agents can reason over consistently.Shared context for operations and AI
Extend the semantic model into the experiences where decisions are made. Ontology-aware Real-Time Dashboards let teams build operational views from entities and relationships rather than raw tables, while Ontology Rules capture business logic and policies that can be interpreted consistently by people and agents. Ontology MCP tools make ontology definitions and queries available to external agents and developer tools, and Data Agents using Ontology as a context source can use mappings, synonyms, relationships, and bindings to produce more accurate and explainable queries. The planned integrations with Foundry IQ, CPS, and Operations Agent extend that shared context across Microsoft AI and operational workflows.
Trusted, governed platform
Apply enterprise controls as the ontology becomes a shared foundation for analytics and AI. OneLake Security and RLS ensure that ontology authoring and querying respect the permissions applied to the underlying data, while management APIs, CI/CD support, and the Ontology Management SDK are intended to support repeatable administration and deployment. Graph data materialization can be enabled selectively for exploration and multi-hop analysis, keeping graph generation an intentional choice rather than a requirement for every workload.
Together, these advancements strengthen the foundation for enterprise intelligence: a shared, continuously updated understanding of the business that connects signals, context, decisions, and actions. As AI becomes embedded in every workflow, the defining question is no longer whether you have AI. It's whether your AI operates from a live, trusted understanding of the business.
What does your AI run on?
AI creates a new competitive boundary. The question is no longer how many agents an organization can deploy. It is whether those agents can reason and act from the same live, trusted understanding of the business.
Can your data platform connect everything the business knows, preserve how entities change over time, operate with real-time freshness, support continuous intelligence at machine scale, govern every decision, and take action across the enterprise without forcing teams to assemble disconnected products?
Figure: Enterprise intelligence requires more than isolated AI capabilities. It depends on a platform that can connect data, preserve context, reason in real time, govern decisions, and drive action across the business.Fabric provides that foundation. It brings data, business context, intelligence, governance, decisions, and actions into one shared platform. As knowledge and outcomes accumulate, every new agent and application can inherit a richer operational understanding.
The race is not simply to deploy AI. It is to build the foundation that makes AI effective.
Trusted AI runs on Fabric.
Learn more
To learn more about the capabilities announced at FabCon and explore how Microsoft Fabric is helping organizations build the foundation for enterprise intelligence, visit the resources below.
- Complete the tutorial: Real-Time Intelligence | Ontology
- Ask questions on the forum: Real-Time Intelligence | IQ
- Submit ideas and vote: Real-Time Intelligence | IQ
- Use the docs: Real-Time Intelligence | IQ
- Complete the learning path: Real-Time Intelligence
- Get certified: Real-Time Intelligence
- Read the blog: Real-Time Intelligence | IQ
- Check the release plan: Real-Time Intelligence | IQ
- Explore the Microsoft Fabric documentation.
- Watch Fabric YouTube.
Microsoft Employee