Fabric Databases— a unified, SaaS-native experience for modern data workloads (Generally Available)
Today marks a major milestone: Fabric databases are now generally available, bringing together the strengths of SQL database and Cosmos DB within Microsoft Fabric. This launch redefines how organizations manage, analyze, and activate their data, creating a unified foundation for the next generation of AI-powered apps and innovation.
Fabric databases: simple, autonomous, optimized for AI
We are at a transformative inflection point. AI is now essential, and success depends on more than just collecting data. It’s about handling more complex data, from operational and analytical, structured and unstructured, all in one place. That’s exactly what Fabric was designed to do: bridge every part of your data estate so teams can unlock real-time insights and drive innovation faster. And with Fabric Databases, building apps is simpler than ever: you can provision in seconds, without deep database expertise or complex configuration.
Fabric databases offer serverless, autonomous architecture, enterprise-grade security, and native AI integration, including support for vector data and RAG patterns, making it the launchpad for modern, data-driven innovation. Developers can query live operational data from SQL database or Cosmos DB without complex ETL or duplicated pipelines, while OneLake keeps the data analytics-ready and in sync for a consistent, real-time view across workloads. Unified security across the platform ensures your operational workloads are protected with enterprise-grade standards, while unified billing simplifies cost management at global scale.
Top Fabric Database Scenarios
Many organizations are achieving operational excellence thanks to Fabric databases. Here are some of the main ways customers are using Fabric databases:
Modern App Development
Fabric databases speed up app development through seamless integration with VS Code, GitHub, and built-in features like GraphQL APIs and User Data Functions. Eastman, a global specialty materials company, is leveraging these integrations in a new agentic application. Before adopting Fabric, their sales teams struggled with fragmented CRM data and time-consuming prep for customer meetings.
Today, they built an AI-powered sales copilot using SQL database in Fabric, Azure Web Apps, and Azure Open AI Service. Sellers simply ask questions in natural language by voice or text. The sales copilot then translates them into SQL queries, and delivers instant answers, dynamic reports, and trend insights. “SQL database in Microsoft Fabric is more precise than basic search,” says Logan Finke, Principal AI Data Architect at Eastman. “It combines vector search with SQL filtering to find the most relevant answers to complex questions.”
Reverse ETL: Real-Time, Actionable Data
Reverse ETL in Fabric transforms analytics from a passive reporting tool into an active driver of business outcomes. Traditionally, data is collected, analyzed, and surfaced in dashboards—leaving business users to interpret insights and take action, often hours or days later. But with Fabric, curated analytics data can be pushed directly back into operational systems like CRM, ERP, or real time personalization services, leveraging the high concurrency of OLTP-optimized databases, low latency and horizontal scale of Cosmos DB, and rich query of SQL DB. For example, as soon as an insight is generated, it can automatically trigger actions—such as sending personalized emails, adjusting inventory, or updating customer records—in real time. The result: analytics becomes actionable, closing the loop between data, insight, and execution, and putting intelligence directly into the systems where decisions are made and actions happen.
For SQL database in Fabric, the Eastman team used shortcuts in OneLake to make the relevant lakehouse and warehouse data available to the application workspace. They then created a pipeline to extract relevant metadata, use AI to preprocess and enrich the data with vector embeddings, and store it back into the CRM. This allowed Eastman to cut development time by eliminating multiple ETL pipelines, and ultimately reduced sellers’ preparation time from 4 hours to 40 minutes. “Our teams feel empowered,” says Andrew Ervin, Generative AI Manager at Eastman. “They are spending less time on admin and more time building relationships. That’s the type of innovation that changes lives, not just workflows.”
For Cosmos DB in Fabric, a well-known e-commerce website, for example, needed to add “You may also like” real-time recommendations experience to customers just before they check out. The recommendations had to be served in under 200ms latency end-to-end to catch users before they click check-out. The personalized recommendation matrix produced in OneLake was stored back to Cosmos DB to meet the low latency and high concurrency demands of the e-commerce website peak traffic.
Translytical task flows
Translytical task flows combine operational and analytical actions in one seamless experience. With Fabric databases and Power BI write-back, organizations can update records, validate inputs, and trigger workflows—all without leaving the report. This capability is a game-changer for businesses that need instant insights and immediate action. “The beauty of Fabric is its translytical approach —combining transactional speed with analytical scale,” says Erik Lidman, CEO at Aimplan, a Microsoft partner. “Our customers can now plan, forecast, and report in the same space, with instant updates flowing across the business.”
AP Pension, a leading Danish pension fund, adopted Fabric to modernize its data delivery. By integrating SQL database in Fabric with a custom Power BI visual, AP Pension enabled real-time write-back and automated governance across 40+ workspaces. The result was faster processing, reduced costs, and improved data quality for analytics and reporting—all within a single capacity.
Kinectify, an anti-money laundering and compliance service for gaming industry, relies on PowerBI as the user experience, enabling operators to invoke actions once anomaly is detected directly from the dashboard, thanks to the Cosmos DB in Fabric integration with User Data Functions and PowerBI visual controls. Mike Calvin, CTO of Kinectify, highlights: “Cosmos DB in Fabric has been a home run for Kinectify. As company we store a lot of data in Cosmos, and being able to automatically, reliably and performantly move data with zero ETL between Cosmos DB and One Lake in Fabric has allowed us to experiment with Fabric in ways that we simply would not have been able to otherwise.”
Getting Started
SQL database in Fabric provides reliable, structured relational databases within Fabric, using the same trusted SQL Server and Azure SQL Database engines. Developers can quickly build scalable, secure, intelligent solutions. Customers like Aon, CSX, and Save the Children, and partners like Accenture/Avanade, Cloud Services, and Hexaware have created more than 50,000 databases during public preview. They are already streamlining operational and analytical workloads for AI-driven apps. Benefits include fast setup, Copilot-powered convenience, T-SQL support, integration with SSMS and VS Code, and built-in CI/CD. New security features include SQL Auditing for greater transparency and Customer Managed Keys for more enhanced control. Watch the Data Exposed series to catch the latest updates:
Cosmos DB in Fabric provides a distributed, low latency, high performance NoSQL database at any scale for your semi-structured within Fabric, using the same trusted Azure Cosmos DB service. With Cosmos DB in Fabric, developers can set up low latency serving layer for One Lake data, work with semi-structured forms data in power BI translytical apps, store features and metadata for their Spark jobs and pipelines, and more. With Cosmos DB vector indexing and search capability, customers can index their One Lake data and serve vector queries over one Lake data in milliseconds. Whether building dashboards, running analytics, or training AI models, teams benefit from Cosmos DB’s low-latency architecture and production-grade performance—all within a single, integrated Fabric experience that accelerates insight and innovation.
The Road Ahead
With the general availability of Fabric databases, Microsoft is delivering a unified, SaaS by default, PaaS configurable database for modern data workloads. But the innovation doesn’t stop here. We are also introducing a new set of data platform capabilities; you can learn more in the blog post on Fabric Databases—bringing together the best of SQL database and Cosmos DB in Fabric.
You will see that every innovation landing in Azure SQL database also comes into SQL database in Fabric—all on the same trusted SQL foundation. Stay tuned as we advance Fabric databases capabilities to further simplify provisioning, accelerate insights, and enable organizations to manage and activate data at scale.
Ready to learn more?
Explore our Fabric databases documentation, tutorials, and videos to see how Fabric databases can power your journey to AI-driven innovation. Register for our upcoming webinar to learn more.
If you’re attending Microsoft Ignite this week, don’t miss the following sessions to see these capabilities in action and hear directly from the product teams:
BRK134: Modern data, modern apps: Innovation with Microsoft Databases
BRK220: SQL database in Fabric: The unified database for AI apps and analytics
BRK228: Real-time analytics and AI apps with Cosmos DB in Fabric
