Trusted AI starts with Microsoft Fabric: Unified real-time intelligence and IQ context
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.
Modern businesses operate in environments where conditions change continuously, and many decisions cannot wait hours. Speed alone does not create alignment. Many platforms focus on moving data faster — through streaming pipelines, dashboards, alerts — but without shared context, teams and AI systems interpret signals differently. Insights fragment. Decisions diverge. Modern businesses need more than faster data. They need a unified operational view. To compete, organizations must operate in real time: sensing what is happening, understanding its context, and responding in time to change outcomes across both digital systems and physical environments.
Yet signals and insights alone are not enough. Modern systems generate overwhelming volumes of events, charts, dashboards, and alerts. Without structure, they fragment understanding instead of improving it. What businesses need is unified context: a live view that connects temporal, spatial, geospatial, and relational data with operational actions across systems, assets, and environments. With this semantic foundation, people and AI agents can reason about what is happening and what to do next in the language of the business.
This shared context is essential for trusted enterprise AI. AI systems cannot reason, decide, or act reliably without a consistent, up-to-date understanding of the environment in which they operate. Fabric transforms fragmented signals into a continuously updated operational view that people and AI agents interpret the same way. This same foundation is what allows AI to operate reliably when decisions affect real-world systems, not just digital workflows.
Operating from this common context keeps teams and AI aligned. AI agents monitor conditions, reason over business entities and relationships, and act autonomously within defined guardrails. People set intent, apply judgment, and provide oversight. Decisions propagate faster and coordinated action becomes possible across complex systems.
Microsoft Fabric provides this real-time operational foundation through Real-Time Intelligence and Fabric IQ. Today, we are also announcing a partnership with NVIDIA that extends this intelligence platform into the physical world. By integrating Fabric with NVIDIA Omniverse, businesses will be able to connect real-time data and business context with 3D visualization and physical simulation, enabling spatial intelligence and the next generation of physical AI.
Microsoft and NVIDIA partner to deliver an integrated platform for the next generation of physical AI
Because Fabric already unifies real-time signals, shared context, and action, extending that loop into physical simulation is a natural next step. Today, we are announcing a partnership between Microsoft and NVIDIA to power the next generation of Physical AI by integrating Microsoft Fabric with NVIDIA Omniverse libraries. Together, the platforms unify real-time operational data, business context, and physical simulation into a single system for understanding and optimizing complex physical environments.
Many operations run in physical environments such as airports, logistics hubs, and factories. Managing them requires more than data. Organizations must understand how events evolve across time, space, and relationships, and how decisions affect the physical world. By combining Fabric with Omniverse, organizations can connect real-time data with spatial context to gain a unified operational view.
As part of this collaboration, we are working with Vanderlande, a global leader in automated systems for airports and logistics hubs. Vanderlande is using Microsoft Fabric and NVIDIA Omniverse to build an AI-powered 3D Airport Digital Twin using Vanderlande’s OpenAir Platform, creating a unified operational view across airside, terminal, and baggage operations.
This solution brings together three complementary layers:
- Fabric Real-Time Intelligence for real-time data streaming and analytics.
- Fabric IQ for shared business context and semantics.
- NVIDIA Omniverse Libraries for 3D visualization and physical simulation.
Together, these layers create a unified system where organizations can sense, understand, simulate, and act. Live signals flow into Fabric, become shared business context, and power immersive digital twins in Omniverse where teams can visualize and understand operations before acting. The same real-time context that drives decisions now informs how physical systems are understood and optimized.
Figure: Microsoft Fabric + NVIDIA – the partnership.
The first step will arrive in private preview in April, enabling teams to embed Omniverse 3D scenes into Fabric Real-Time Dashboards. Bidirectional cross-highlighting and filtering will link dashboards and 3D environments, so analytics, visualization, and simulation operate from the same real-time data and context.
Microsoft Fabric: the operational foundation
Fabric is designed for how modern businesses operate: continuously sensing, analyzing, deciding, and acting across systems in real time. It provides a complete intelligence platform that unifies data, context, and action into a single operational loop for running both digital and physical operations.
Figure: The operational foundation for the modern enterprise: IQ, Real-Time Intelligence, and OneLake.
Many platforms focus on moving data faster through streaming pipelines, dashboards, and alerts. But speed alone does not create alignment. Without a shared semantic foundation, teams and systems interpret signals differently. Insights remain fragmented, decisions diverge, and AI agents struggle to earn trust. In real-world operations, misalignment doesn’t just slow decisions, it directly affects safety, efficiency, and outcomes.
Fabric takes a different approach. It unifies the full intelligence stack: ingesting signals, transforming them into rich context, reasoning over that context, and enabling coordinated action. Real-time signals connect to business entities through a shared ontology, forming a live operational view understood by people and AI agents alike.
Fabric delivers this through four foundational layers working together: OneLake, Real-Time Intelligence, IQ, and AI agents.
OneLake: unified data foundation
OneLake is the unified data lake at the core of Fabric, bringing together real-time operational signals and analytical data into a single data estate. With mirroring, cross-cloud shortcuts, and hundreds of connectors, OneLake eliminates silos that slow decisions and fragment understanding.
When data lives in one place, teams no longer reconcile competing versions of reality or wait for pipelines to sync. People, applications, and AI agents operate from the same trusted data foundation.
Figure: Data from OneLake eliminates silos and brings data to a single estate.
Real-Time Intelligence: unified time, space, and relationships
Real-Time Intelligence powers Fabric’s operational capabilities through a continuous lifecycle: Stream → Analyze → Model → Visualize → Act, turning signals into operational decisions. It unifies three key dimensions that are fundamental to businesses and their operations: time, space, and relationships, the same primitives needed to model how the real world behaves.
Figure: The main components of Real-Time Intelligence.
These capabilities span the real-time lifecycle:
- Real-Time Hub: Discover and manage streaming data in one place.
- Eventstream: Ingest events through an extensive connector ecosystem.
- Eventhouse: Store and analyze time-series data at cloud scale.
- Real-Time Dashboards: Visualize live operational data with sub-second refresh.
- Activator: Detect patterns and trigger automated responses.
- Anomaly Detector: Identify unusual patterns using machine learning.
Together, these capabilities collapse the time between signal and action, so teams can detect issues earlier and change outcomes while there’s still time. In physical operations, that time gap is often the difference between smooth flow and costly disruption.
Fabric IQ: unified intelligence and business context
Operating in real time requires shared context. Teams must understand what data represents, how entities relate, and what actions are possible across the business. Fabric IQ provides that context through a shared ontology. It organizes data in the language of the business by defining core entities such as customers, locations, assets, transactions, and teams, and binding them to both historical and real-time data. This semantic layer turns raw signals into business meaning that people, applications, and AI agents can interpret and act consistently in both digital workflows and physical operations.
Hanwha Qcells is applying this model to data center energy management.
“The scale and complexity of energy management in modern AI data centers now exceeds what humans alone can operate, requiring AI-driven, autonomous energy orchestration. Hanwha Qcells is using Fabric IQ to add agentic AI to our Energy Management System, enabling seamless human-agent collaboration. At the core of this is a data center ontology built on Microsoft Fabric, which creates a shared language that allows human operators, data teams, and AI agents to work toward the same goals. This shared context enables AI agents to deliver actionable recommendations while keeping final decisions with the operator, driving cost savings, resilience, compliance, and emissions reduction.”
– Emmanual Daniel, VP, Solutions Department.
This is what trusted AI looks like in practice. Because dashboards, analytics, and agents operate on the same model, the organization shares a common understanding of how the business, and its operations function. This eliminates fragmented interpretations and enables faster coordination and execution.
Figure: Microsoft IQ unifies the act–decide–analyze–observe loop, creating a shared semantic foundation that keeps people and AI aligned in real time.
Fabric IQ is part of a broader Microsoft IQ vision that unifies operational context with the knowledge and flow of work across the organization:
- Work IQ: Communications, documents, and the flow of work.
- Fabric IQ: Live real-time state and operational actions of your business.
- Foundry IQ: Curated institutional knowledge.
AI agents: intelligent reasoning and autonomous action
AI agents in Fabric operate over the same shared context as people. They monitor live signals, detect patterns, surface recommendations, and act within guardrails defined by humans; this is especially critical when actions affect physical systems where safety, compliance, and reliability are non-negotiable. Fabric includes two types of agents:
- Data agents provide conversational access to the business. Teams ask questions in natural language and receive answers grounded in the ontology and real-time state of the business.
- Operations agents continuously monitor conditions, identify patterns, and take autonomous actions within defined boundaries. Humans define intent and accountability. AI executes at machine speed.
They form a continuous intelligence loop: data flows in, agents reason over shared context, insights emerge, and actions flow back into the business. Each action generates new data that refines the model, enabling the system to improve over time.
What’s new in Fabric
Fabric continues to evolve rapidly, with major new capabilities reaching general availability and new innovations entering preview. These are the most significant announcements:
Maps (Generally Available)
Figure: Maps provide a rich geospatial layer to the data in Fabric.
Maps are now generally available in Fabric, bringing native geospatial intelligence directly into your real-time operating model. Maps make location a first-class dimension of operational decision-making, so where something is happening becomes just as important as when it’s happening and how it’s connected across your business.
Integrated with Fabric’s real-time analytics and shared ontology, Maps enable teams and AI agents to visualize and act on spatial context in real time, powering scenarios like asset tracking, facility monitoring, and logistics optimization without moving data outside Fabric.
For a deeper look at Maps capabilities, read the Maps in Microsoft Fabric blog.
Business Events
Business Events provide event-driven architecture to Fabric at the semantic level. Instead of monitoring raw telemetry or database changes, you can define, detect, and act on business-level occurrences—things like “customer at risk,” “shipment delayed,” or “equipment needs maintenance.”
Business Events work with your ontology to detect meaningful patterns across streaming and static data. When conditions match, Fabric raises a business event that can trigger workflows, activate alerts, or invoke AI agents. This shifts monitoring from technical signals to business outcomes.
Business Events closes the gap between technical systems and business understanding. To go deeper on how Business Events work and the full set of capabilities, read the Business Events blog.
Graph improvements
Fabric Graph continues to expand, adding support for billionsscale graphs, deeper Fabric Data Agent integration, expanded GQL capabilities including shortestpath queries, and continuous UX improvements that simplify modeling, querying, and exploration for both humans and AI agents.
Figure: Query across relationships using a unified graph to uncover patterns and insights that span customers, locations, and assets.
Graph works directly on data stored in OneLake, eliminating the need to copy or move data into specialized graph databases. You define relationships in your ontology, and Graph makes those relationships queryable as live conditions change. This allows teams and AI agents to understand not just what happened, but how changes propagate across the business in real time.
More to come at the end of April.
Ontology improvements
The ontology capabilities in Fabric IQ continue to expand with several significant improvements focusing on operating a shared business model at scale across teams, agents, and workflows without losing trust, clarity, or control. These updates strengthen how real-time signals connect to business entities, embed rules and actions directly into operational context, and apply enterprise-grade security and governance as the foundation for AI in Microsoft Fabric.
Learn more in the full Fabric IQ Ontology blog.
SQL Operator in Eventstream
SQL Operator brings familiar SQL into real-time stream processing in Eventstream, making it easier for teams to turn live signals into action. Data engineers can apply precise SQL logic over streaming data to detect conditions, filter noise, and drive low-latency outcomes at scale. By lowering the barrier between insight and execution, SQL Operator helps organizations move faster from real-time data to real-time decisions, reinforcing Fabric’s shift from analytics to operations.
Building on that foundation, SQL Operator is evolving to enable content-based routing for real-time operations in Fabric. Teams can now define transformation and routing logic once in SQL and seamlessly fan out live results to multiple destinations such as Eventhouse, Lakehouse, Activator, or downstream streams, without duplicating pipelines or maintaining parallel logic. With built-in testing and a single, maintainable SQL surface, these new capabilities make it easier to design, validate, and operate sophisticated real-time workflows at scale, turning streaming data into coordinated operational action. These capabilities will be available in early April.
For more details on the current capabilities, please see the Fabric Eventstream SQL Operator blog.
Additional updates
The roadmap for Real-Time Intelligence and IQ includes many other enhancements currently in development or preview:
- Enhanced anomaly detection models with multivariate support.
- Improved Real-Time Dashboard performance and interactivity.
- Expanded connector ecosystem for Eventstream.
- Operations agent enhancements for more complex decision patterns.
- Deeper integration between Data agent and Graph queries.
- Additional geospatial analytics in Map.
We’re continuously acting on customer feedback, delivering new fixes and features weekly. The pace of innovation creates a compounding advantage for early adopters who integrate these capabilities into their operations now.
The future is already here
The capabilities described above are available today, and organizations are already using them to transform how they operate. Real-time intelligence is not just about faster data. It is about shared business context, coordinated human and AI action, and decisions made while outcomes can still change, across digital systems and physical operations.
Innovation across Fabric, Real-Time Intelligence, and IQ continues to accelerate, with improvements shipping continuously based on customer feedback. Early adopters gain a compounding advantage. The longer an organization operates with real-time intelligence, unified context, and AI collaboration, the more it learns, adapts, and improves. Meanwhile, competitors still managing operations in cycles fall further behind.
The modern business is already here. The question is whether yours is ready.
Get started today
Ready to unify your intelligence platform? Getting started doesn’t require rebuilding your systems. It starts by defining shared business entities, connecting live signals, and letting Fabric coordinate insight and action across teams and AI agents—incrementally, safely, and in real time. Here’s how to begin:
- 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
- Follow on LinkedIn
- Watch on YouTube
