Connecting apps, databases, and AI on one foundation: new innovations at SQLCon and FabCon EU 2026

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.

AI is changing what organizations need from their databases. The systems that power day-to-day operations can no longer remain disconnected from analytics and AI. Developers and data teams need a trusted foundation that brings live business data, insight, and intelligence together without adding more data movement or operational complexity.

At SQLCon and FabCon Barcelona, we're introducing the next steps in Microsoft's database strategy. Together, the latest innovations in Fabric help organizations move from building applications to managing database estates and operating them more intelligently with AI.

In this post, I'll cover:

• Major new capabilities for application development in Fabric.
• Updates to Database Hub, a centralized experience for managing databases across Fabric.
• Introduction of database agents, an AI-powered approach to database operations.
• New SQL innovations that extend the same foundation across customer-controlled infrastructure and the cloud.

Build intelligent applications directly on enterprise data with Fabric Apps

AI can accelerate prototyping, but moving an application into production still requires teams to connect data, manage infrastructure, and enforce governance. Fabric Apps help bridge that gap, enabling developers to build AI-powered business applications, operational workflows, and custom experiences directly on governed data in Microsoft Fabric.

Since the initial launch in June, thousands of developers have used Fabric Apps to turn enterprise data into applications and experiences. Using the open-source Rayfin SDK and CLI, developers and coding agents can build application backends with databases with built-in authentication, then deploy them to Fabric as Fabric Apps.

New capabilities that make Fabric Apps more flexible and production-ready are:

• Data connectors: Connect your application to existing data in Fabric warehouses, Lakehouse's, SQL databases, and semantic models.
• PostgreSQL support: You can now create and use a PostgreSQL database as your application’s transactional store.
• Flexible, policy-controlled access: Keep applications private by default, apply tenant-wide hosting and access policies, and support approved public-facing experiences such as product catalogs, event portals, and status pages.
• Backend functions: Extend your application with TypeScript functions that run business logic, process data, and connect to external services, with Secret Store protecting sensitive credentials.
• Integrated application storage (coming soon): Manage files alongside relational data with OneLake-integrated storage defined in code and deployed through the same application lifecycle.

Together, these capabilities give developers a faster path from idea to a secure, governed, production-ready app in Fabric. Explore the Fabric Apps blog to learn more.

Figure: An AI-powered travel planning app brings planning, personalized recommendations, and expense management together in one place, enabling users to create and adjust itineraries and track spending for themselves or their team.

Manage your entire database estate from a single experience with Database Hub

As organizations build more applications and connect them to more data, database operations become increasingly complex. Many database admins and platform teams manage environments using multiple tools, portals, and monitoring systems. Visibility becomes fragmented, governance becomes more difficult to maintain consistently, and opportunities for optimization are often discovered only after issues emerge.

Database Hub in Fabric addresses this challenge by providing a unified experience for exploring, observing, governing, and optimizing databases. Now in preview, Database Hub provides a centralized control plane for managing SQL Server, Azure SQL, PostgreSQL, Cosmos DB, and SQL database in Fabric. Teams can understand the health, performance, security posture, and optimization opportunities across their database estate.

Instead of viewing databases in isolation, admins gain broader visibility into the overall environment, helping them identify emerging bottlenecks, prioritize operational work, and apply governance practices more consistently.

The goal is to help database teams spend less time navigating management complexity and more time improving the systems that power the business.

Figure: A sample overview page of the Database Hub that allows users to monitor the overall health of their data estate.

From monitoring to intelligent operations with database agents

With database estates growing, organizations generate more signals than ever before. Performance metrics, security findings, health indicators, availability alerts, and cost insights continuously stream into operational workflows.

For years, database operations have largely been reactive. Teams monitor dashboards, investigate alerts, correlate signals, and determine the appropriate response. As environments become larger and more complex, this operating mode becomes increasingly difficult to scale.

Database agents for SQL and PostgreSQL on Azure represent an important step toward AI-assisted database operations. Now rolling out in preview, these agents are accessible in the Database Hub in Fabric and VS Code. They continuously analyze performance and workload health, helping teams identify issues faster and focus on the actions that matter most. They help prioritize recommendations, accelerate investigation, and support faster resolution of operational challenges. SQL database agent combines AI model reasoning with SQL Database’s native intelligence including Intelligent Query Processing and a domain-specific development and evaluation harness informed by decades of database engineering expertise.

Human oversight remains central to the experience. Agent actions operate within role-based access controls, approval workflows, observability, auditing, and operational guardrails, helping organizations maintain governance and accountability while benefiting from AI-assisted workflows.

Figure: Database agent settings, where administrators can enable the agent and select the databases it monitors.

Beyond operations, database agents also help organizations create a stronger foundation for AI. By exposing the agents’ capabilities through MCP and skills and by leveraging Fabric's connection to operational data through OneLake, they help make trusted business information more accessible to applications and intelligent agents.

The result is a shorter path from operational data to grounded AI experiences, with less integration work, fewer custom pipelines, and greater productivity across teams.

Extending innovation across Fabric and SQL

Fabric brings applications, analytics, database management, and AI together in a unified experience. The broader Microsoft SQL portfolio extends that innovation across environments, helping organizations keep critical workloads close to their operations or scale them in the cloud without compromising performance or consistency.

• For organizations with sovereignty requirements, maintaining control over sensitive data, where it resides, and how critical systems operate is essential. SQL Server on Azure Local is now generally available, bringing trusted SQL Server capabilities to Azure infrastructure for low-latency, high-performance workloads. SQL Server on Azure Local Disconnected Operations is also generally available, bringing these capabilities to sovereign and isolated environments without continuous cloud connectivity.
• Azure SQL Database Hyperscale continues to advance performance, scale, and efficiency for mission-critical applications. New capabilities include serverless auto-pause for intermittent workloads, which stops compute billing when databases are paused, and higher-density elastic pools supporting up to 50 databases, helping customers consolidate resources and reduce cost per database. Beyond that, other generally available releases include 192-vCore compute to support demanding workloads, and DiskANN vector indexing is now generally available, combining vector search with T-SQL so developers can build RAG and recommendation experiences without moving data. SQL’s vector index now has full DML support allowing you to modify your existing table with vectors and the index is automatically updated.

Together, these innovations give organizations the flexibility to run mission-critical workloads where they perform best, scale without rearchitecting, and build AI experiences on a consistent SQL foundation. Read more about the new SQL-related capabilities.

Get started

Across Fabric Apps, Database Hub, database agents, Fabric, and SQL, the goal remains the same: to help organizations connect applications, data, operations, and intelligence into a single, coherent platform experience.

As AI becomes an integral part of every application and business process, the organizations that succeed will be those that can move seamlessly from data to insight to action. The innovations we're announcing at FabCon and SQLCon Barcelona represent another important step toward that future, helping customers build, manage, and scale intelligent applications with confidence.

I’d like to share that we are publishing our SQL roadmap which shows the upcoming features which will be released.

If you are attending FabCon and SQLCon in person, I encourage you to:

• Attend the database sessions, from deepdive workshops on Fabric databases to SQL keynotes on the future of databases in an AI driven world. Check out the full agenda.
• Connect with experts — our database engineers and product leaders are onsite and eager to hear your feedback. Come visit us at the Ask the Experts booths in the expo hall.
• Take the new SQL AI Developer Certification Exam and Azure Cosmos DB AI Developer Certification to boost your skills with AI.