Last week at FABCON & SQLCON - The Microsoft Fabric & SQL Community Conferences 2026 in Atlanta, Microsoft delivered a wave of important announcements for Microsoft Fabric.

From major GA milestones and long-awaited integrations to meaningful progress in governance and AI—the momentum is undeniable.

But if I had to pick one announcement that truly stood out, it would be Planning in Fabric IQ. The fact that Satya Nadella highlighted it during his keynote, reinforced by Arun Ulag, and with Amir Netz dedicating meaningful time to it, says a lot about its importance. The message is clear: integrated, enterprise planning and forecasting is now a native capability within Microsoft Fabric.

As a result, agents developed by organizations can move beyond answering questions about historical data—they can now reason about the future, leveraging forecasted data to provide more forward-looking, actionable insights.

For me, this doesn't mean that other announcements aren’t significant—they absolutely are—but because this one fundamentally changes how organizations move from insight to action.

If you want a deeper dive into Fabric IQ, I recommend checking out my previous article: “Why Organizations Want Fabric IQ?”

Why Planning in Fabric IQ matters

For years, analytics and planning have lived in separate worlds.

We analyze what happened in Microsoft Fabric… …and then plan the future somewhere else.

Different tools. Duplicated logic. Exports, reconciliation, and inevitable drift. Planning in Fabric IQ finally closes that loop.

Planning now sits directly on top of:

• Governed Fabric data

• Shared semantic models (where we can prepare our data for AI)

• OneLake as the unified data foundation and using open source formats such as Delta Parquet and iceberg to store the data

Budgets, forecasts, targets, and what-if scenarios are no longer downstream artifacts—they become a native extension of analytics.

This represents a fundamental shift—from reporting on the past to actively shaping the future—all built on the same semantic and governance layer.

Why this is critical for AI

For anyone serious about agentic AI, this is even more important.

AI doesn’t just need to understand what happened. The added-value increase when business users can understand the intent. Planning is where intent lives.

By embedding planning into the same foundation as analytics, Microsoft Fabric enables AI to reason across Actuals, Plans and Scenarios.

All in a unified, governed context.

And yes—there were many other big announcements

FABCON2026 delivered several major milestones. Since I work with large enterprise organizations, below you can see the one I like the most:

• Mirroring for Oracle and SAP Datasphere is now GA, reducing ETL complexity and latency for critical enterprise systems

• Shortcut AI transformations reaching GA make OneLake reuse smarter—not just more cost-efficient

• Continued evolution of OneLake security, strengthening Fabric as an enterprise-grade data control plane

• Fabric Data Agents now GA, pushing Fabric further toward action—not just insight

• Fabric MCP Local is now GA and remote MCP entered public preview, letting Al agents like GitHub Copilot act directly inside Fabric.

• New Git integration capabilities in (Preview)that for sure, will improve the developer experience in Fabric (Branched Workspace, Selective Branching & Compare Code Changes)

All of these make Fabric more open, more governed, and more AI-ready.

But Fabric IQ Planning is what ties it all together

What makes Planning in Fabric IQ unique is how it connects the full lifecycle:

It transforms Fabric from a platform that explains the business into one that actively helps steer it.

That’s why, among all the announcements this week, this is the one I’m most excited about.

The future of data platforms isn’t just about combining analytics and AI. It’s about going a step further—bringing together analytics, intent, and action on a single, governed foundation.

Speaking of future, the roadmap includes Autonomous Planning powered by Ontology Models and Data Agents all using Fabric Capacity Units (CUs), the same consumption model as the rest of Fabric. You can learn about CUs and cost management by reading the edition #7 of my newsletter: The Economics of Modern Data Platforms (Microsoft Fabric vs. Azure Databricks) | The Data Massagist at Linkedin

You can learn more about Fabric IQ Planning by reading Dipti Borkar's post: Introducing Planning in Microsoft Fabric IQ: From historical data to forecasting the future

A quick look at the architecture behind Planning in Fabric IQ

What makes this even more powerful is the simplicity of the architecture. There is no new silo, no parallel planning system, and no need for custom synchronization layers.

The real leap forward is that this architecture enables Fabric IQ to go beyond answering what happened and what is happening today. It also allows organizations to understand what should happen next—and, just as importantly, why. That’s where the real transformation lies.

At a high level:

• Source systems (Oracle, SAP, SQL, SaaS, ...) land in OneLake via mirroring, shortcuts, or pipelines of Data Factory,

• Fabric semantic models define business logic once

• Fabric IQ Planning sits directly on top of those models

• Writeback flows into Fabric SQL tables, fully governed

• Data Agents operate across actuals, plans, and scenarios in the same context

In simple terms: Same data. Same semantics. Same governance.

Because everything lives inside Microsoft Fabric:

• No data duplication for planning

• Security, lineage, and access control remain consistent

• BI, AI, and operational workloads share the same version of truth

• Plans don’t drift—they evolve with actuals

Fabric IQ Planning brings together four integrated components that support the full business workflow—from data integration to planning, reporting, and analytics.

• Planning sheets: No-code, collaborative workspace for budgeting, forecasting, and operational planning

• PowerTable sheets: Enterprise-grade reporting and structured data management with advanced analysis capabilities

• Intelligence sheets: Self-service analytics with dashboards, visualizations, and KPI tracking

• InfoBridge: Unified data integration layer that connects, transforms, and prepares data across sources

I cannot wait to try it and work with a customer on a real scenario.

Understanding “SQL Write-Back” in Planning

One concept that wasn’t immediately obvious to me when Planning in Fabric IQ was introduced was SQL write-back—but it’s actually one of the most powerful aspects of the architecture.

Planning in Fabric IQ doesn’t rely on a separate planning engine or proprietary store. Instead, all planning inputs—budgets, forecasts, scenarios, and adjustments—are written directly into Fabric SQL databases.

In practice, this means:

• Users enter planning values directly in Fabric IQ

• Those values have persisted as rows in Fabric SQL tables

• Data is stored in a relational, governed, and fully queryable format

• Plans can be seamlessly joined with actuals using standard T-SQL

• Power BI, notebooks, pipelines, and data agents could see updates instantly

There’s no export. No synchronization. No duplication.

This works because Fabric SQL Database provides the foundation needed for enterprise-grade planning:

• Transactional writes for consistency

• Schema control for scenarios, versions, and keys

• Built-in security and governance

• Predictable performance for frequent updates

As a result, planning data becomes first-class enterprise data—living alongside operational and analytical data, instead of drifting outside the platform.

If you want to try it (and I’m sure you will), make sure the Fabric tenant setting “Users can create Planning (preview) items” is enabled, as it’s required to create plans in Microsoft Fabric. You can learn more in the official Microsoft documentation: What is Fabric IQ Plan (preview)on Microsoft Learn

I'm curious to hear your perspective: Which #FABCON2026 announcement do you think will have the biggest real-world impact?

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 FabCon and SQLCon announcements across both Microsoft Fabric and Microsoft database offerings.