Advancing Databases for the Next Generation of Applications
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.
New announcements from FabCon | SQLCon
As organizations accelerate their AI adoption, the role of databases is fundamentally changing. It’s no longer enough to manage operational and analytical data separately. AI-driven applications demand a unified foundation where transactional and analytical data come together securely, in near real time, and at scale.
At FabCon and SQLCon, we are sharing how Microsoft is reimagining databases to meet this moment. With a single SQL engine spanning on-premises, PaaS, and SaaS—and with databases built natively into Microsoft Fabric—we’re helping customers modernize SQL, unify their data estate, and build AI-native applications faster and with greater confidence.
In this post, I’ll walk through the latest database innovations in Fabric that continue to advance this vision: bringing greater simplicity, unification, and enterprise readiness to Microsoft databases.
Introducing the Database Hub in Fabric: An agentic control plane for your database estate
As data estates grow, so does operational complexity. Most organizations manage a mix of relational and NoSQL databases across edge, PaaS, and SaaS environments, often through fragmented tools, portals, and management experiences. As AI places even greater demands on data estates, unifying databases under a single access point and control plane has become essential.
Microsoft Fabric is expanding its role as the central access point for enterprise data to help address this challenge. We are introducing the Database Hub in Fabric, now available in early access, as a unified database management experience that brings together databases across edge, PaaS, and Fabric into a single, coherent view. Database teams now have one place to explore, observe, govern, and optimize their entire database estate, including Azure SQL, Azure Cosmos DB, Azure Database for PostgreSQL, SQL Server (enabled by Azure Arc), Azure Database for MySQL, and Fabric databases, without changing how each service is deployed.
Figure 1: The Database Hub in Fabric is a new unified database management experience that brings together all databases into one coherent view
Built for scale, the Database Hub introduces an agent-assisted, human-in-the-loop approach to database management. Intelligent agents continuously reason over estate wide signals to surface what changed, explain why it matters, and guide teams towards what to do next. Built-in observability, delegated governance, and Copilot powered insights help teams quickly understand what’s happening across their database estate and why. Aggregate health views, common performance categories, and trend analysis provide consistent signals across services, enabling operations and development teams to move from insight to action with greater confidence.
Over time, this agent assisted model lays the foundation for increasing autonomy, where database estates don’t just surface issues faster, but continuously learn, adapt, and improve how they’re managed, while humans define goals, boundaries, and trust, and agents provide the scale to make that intelligence compound.
With the Database Hub, teams spend less time navigating tools and more time enabling what comes next: unlocking deeper integration for building applications, analytics, and AI, all from a single control plane for the Microsoft databases portfolio.
SQL database in Fabric: Modernizing SQL for an AI-driven world
SQL databases remain foundational to modern applications. In Fabric, they are designed for a new era. SQL database in Fabric brings the same Microsoft SQL engine into a SaaS-native, autonomous experience, extending Azure SQL Database so teams can connect operational data to analytics and AI when they’re ready.
Customers such as Origence, a fintech platform serving credit unions, show what this looks like in practice. They modernized its legacy SQL Server environment by moving to SQL database in Fabric and simplifying ETL with OneLake, enabling real-time analytics, lowering infrastructure costs, and preparing AI workloads.
“SQL database in Fabric gives us a modern SQL foundation for high-volume operational data and real-time analytics—helping us move faster without compromising enterprise‑grade security or control.”
— Jeff Shood, CTO, Origence
Today, we’re announcing a new slate of capabilities that make it easier to migrate, manage, secure, and optimize SQL workloads for analytics and AI.
Modernize with confidence: Migration Assistant (Preview)
Modernization is often less about technology and more about confidence. Customers want a clear, guided path to bring existing SQL workloads into Fabric without guesswork or disruption.
To address this, we are introducing the Migration Assistant for SQL databases (Preview). The Migration Assistant helps teams assess readiness, identify compatibility considerations, and guide schema migration from SQL Server into Fabric. This reduces manual effort and uncertainty along the way, making it easier for customers to modernize SQL workloads while preserving existing investments and skills.
Figure 2: Migration Assistant helps customers modernize SQL workloads and migrate to Fabric.
Autonomous by default, configurable where it matters
SQL database in Fabric is autonomous by default: handling scaling, management, and optimization automatically. With new configuration options and enterprise-grade safeguards, customers gain more control without reverting to infrastructure management.
Recent enhancements include:
- Expanded enterprise security controls: Support for SQL auditing, customer-managed keys for encryption control, dynamic data masking, and enhanced availability zone support strengthens security and compliance readiness.
- Expanded SQL compatibility: Support for more database compatibility levels, T-SQL features, full-text search, and ALTER DATABASE settings eases application migration without code changes.
- Full collation support enabled: All Azure SQL collations are now usable in Fabric, enhancing global data handling and app development flexibility. Collations can be specified at database creation via REST API.
- Configurable compute limits: Users can now set database-level compute caps to control vCore usage, aiding cost and scaling management on shared capacities while maintaining safe defaults.
Figure 3: SQL database in Fabric now supports configurable compute limits.
- Improved backup and delivery: Deleted databases enter a soft-deleted state in a Recycle Bin with configurable backup retention from 1 to 35 days, allowing restoration to any point within the retention period.
- Pre/post-deployment script support: Source control and deployment pipelines now execute additional T-SQL scripts for enhanced branching, provisioning, and static data management, compatible with CI/CD tooling.
Together, these capabilities help organizations meet enterprise requirements while keeping operations simple and SaaS native.
Built for analytics and AI
What differentiates SQL in Fabric is not just how it runs, but how it connects. With built-in mirroring to OneLake, SQL data becomes immediately analytics-ready, enabling scenarios such as real-time insights, vector search, and AI-powered applications without custom ETL or duplicated pipelines.
New features include:
- Finer control over mirroring to OneLake: Users can selectively choose which tables are mirrored to OneLake and manage mirroring operations via REST API, improving flexibility for analytics and AI scenarios.
- Enhanced vector indexing for AI workloads: DiskANN Vector Indexes now support advanced quantization, full DML operations, and iterative filtering, improving vector search performance and enabling real-time index maintenance.
These capabilities ensure SQL data doesn’t live in isolation but actively powers end-to-end analytics and AI workflows across Fabric.
Refer to the SQL database in Fabric documentation to learn more.
Cosmos DB in Fabric updates
As organizations increasingly connect operational data with analytics and AI, secure connectivity and network isolation are essential, especially for enterprise and regulated environments.
Supporting private network configurations for Azure Cosmos DB mirroring in Fabric (Generally Available)
Today, we are announcing the general availability of private network support for Azure Cosmos DB mirroring in Microsoft Fabric. With this release, customers can mirror data from Azure Cosmos DB accounts secured with Private Endpoints or VNETs into OneLake.
This capability allows customers to maintain consistent network security and compliance controls while enabling near real-time analytics and AI workloads in Fabric. By supporting private network configurations, Cosmos DB mirroring further strengthens Fabric as an enterprise-ready analytics platform by design.
Figure 4: Mirroring data from Azure Cosmos DB accounts secured with Private Endpoints or VNETs into OneLake.
Extending AI coding assistants with the Azure Cosmos DB Agent Kit
Today it is difficult to imagine application development without coding agents. But what if your agent could be an expert in Cosmos DB data modelling, partitioning, and building resilient, distributed apps? Now it can.
The Azure Cosmos DB Agent Kit is an open-source collection of agent skills that extends AI coding assistants with production-ready Cosmos DB best practices. The kit works with GitHub Copilot CLI, Claude Code, and any other agent skills-compatible tool, turning your AI pair programmer into an instant expert in Cosmos DB data modelling patterns, optimization techniques, and best practices. You can simply install the Agent Kit with a single command:
npx skills add AzureCosmosDB/cosmosdb-agent-kit
Read the Azure Cosmos DB Agent Kit documentation to learn more.
Bringing it all together at FabCon and SQLCon Atlanta
Beyond Fabric, we are continuing to advance SQL across on-premises and the cloud. We deliver consistent, AI-ready experience for modern application development. New capabilities include:
- SQL Server 2025 innovations: New semantic and full text search for AI scenarios, native JSON and REST APIs, change event streaming, and improved security, performance, and reliability.
- Azure SQL Database Hyperscale delivers elastic, resilient scale with strong price‑performance—powering large workloads and AI apps through independent read scaling, higher vCore options, vector search, and secure AI integration.
- Simplified SQL modernization: Azure Arc enabled SQL Server migration delivers built-in discovery, readiness assessments, and guided modernization to Azure.
- Developer productivity: GitHub Copilot in SSMS 22 is now generally available.
These announcements reinforce a clear direction for databases in Microsoft. With one unified SQL engine, customers can modernize existing workloads without friction. With SQL and NoSQL databases deeply integrated into Fabric, operational data becomes immediately available for analytics, vector search, and AI. And with the Database Hub, teams gain a single, intelligent control plane to understand, govern, and optimize their entire database estate from edge to cloud to SaaS.
This is how Microsoft is closing the gap between transactions and insights. By unifying databases, analytics, and AI on one platform, data doesn’t just sit in systems; it actively powers the next generation of intelligent applications.
To learn more about these innovations across Azure Databases announced at FabCon and SQLCon, Azure Databases update.
If you are attending FabCon and SQLCon in person, I encourage you to:
- Attend the database sessions — from deep‑dive workshops on Fabric databases to SQL keynotes on the future of databases in an AI‑driven world. Check out the agenda.
- Try the new capabilities — including the Database Hub (early access), Migration Assistant for SQL (preview), and enterprise ready Cosmos DB mirroring with Private Link and VNET.
- 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 Certification Exam — for all SQL professionals to boost your skills with SQL and AI.
Finally, FabCon and SQLCon are heading to Europe! Join the global data and SQL community from 28 Sep – 1 October 2026 in Barcelona, Spain for hands-on learning, expert insights, and real-world stories.
Register to be a part of it. I can’t wait to see you there.
