The keynote from this year’s Microsoft Ignite is over, and we got a lot of new releases in the space of Data & AI.
First of all, the preview of SQL database directly in Fabric.
This new capability brings all the well-known features from the SQL database directly to Microsoft Fabric. SQL databases have been updated in the last period of time to support AI features like vector storage and vector search and the Copilot addition to both SQL Server Management Studio 22 (announced at PASS Data Community Summit – direct download link: Install SQL Server Management Studio 21 Preview – SQL Server Management Studio (SSMS) | Microsoft Learn).
The use cases for a SQL database in Fabric are many. The first, also linked to the capability of vectors above, is to support the development and deployment of intelligent applications faster. Applications that have a need for small and many transactions, which is not good for the parquet file format.
From my own perspective – this could be the first step to introducing a solution to support reference data and other master records hosted and maintained directly in Fabric. We still need the frontend to support the process, but we now have the possibility to
The data from the SQL databases in Fabric is automatically replicated to OneLake to be used for analytical purposes.
The next version of SQL Server has been released to private preview.
This edition will bring the cloud and on-premise world even closer together with built-in capabilities for AI-models and vector natively in the engine.
Security will be even stronger with direct connection to Entra Id and managed identities. A huge improvement to the built-in query optimization capabilities and an easy to use fail-over reliability. Want more? Then there are also the completely new optimized locking features for SQL Server 2025.
Live mirror of the data from SQL server 2025 directly to Fabric OneLake is a part of the new release, alongside the integrated Azure Arc connectivity.
The Real-Time Intelligence offering in Fabric is now GA – this is a huge leap for the capabilities to work with streaming data and get insights to data in near-real-time scenarios.
With the Eventstream to ingest and process data, the Eventhouse to store and query data, and Real-Time Dashboards to visualize and analyze the data.
Actions based on data trends and outliers are also available with the rules functionality.
New connectors for Eventstream were also announced – Google pub/sub and Azure Event Hub, to mention a few.
Fabric OneLake is now the go-to service for utilizing the storage and readings from all your data estates. Snowflake and DataBricks can now be read directly from OneLake without moving data, and the preview of Open Mirroring features enables companies to read even more data directly from the OneLake service offering.
No data without AI anymore
AI is also a big part of the Ignite keynote this year. We now have even more AI capabilities in Fabric. With the release of AI functions directly in Fabric, which gives everyone the possibility to perform text-related tasks with zero or very little code. In Notebooks we can now, among a huge list of elements, summarize, translate and to sentiment analysis with ease.
The AI skills in Fabric can now be leveraged using the Azure AI Agent services to, e.g., seamlessly join the AI Skills from Fabric with SharePoint.
This year’s Ignite keynote is yet again with a great bundle of new releases. But this is not everything. We will get a deep dive into each area of Microsoft at the upcoming product group sessions.
If you want to hear more about all these new features, don’t hesitate to reach out to us.
“With the additions and releases to Microsoft Fabric and the preview release of SQL Server 2025, we see a new era in data and analytics from Microsoft.
Fabric is now even more mature and ready to directly compete with other vendors in the market, and the GA of Real-Time Intelligence sets the standards for the way of working with streaming data”
Brian Bønk, Senior Principal at Intellishore.