Microsoft Databases and Microsoft Fabric: Your unified and AI-powered data estate

4 months ago 101

Today, I’m thrilled to denote the adjacent procreation of Microsoft’s databases: SQL Server 2025, Azure Document DB, Azure Horizon DB, and Fabric Databases, each redesigned to conscionable the demands of AI.

As AI reshapes each industry, 1 information remains constant: information is nary longer conscionable an asset—it’s your competitory edge. The gait of AI demands casual information access, faster insights, and the quality to iterate without friction. Yet galore organizations are held backmost by fragmented information estates and bequest systems. Microsoft Fabric was designed to conscionable this moment—to unify your data, simplify your architecture, and accelerate your way to becoming an AI-led organization.

That ngo is gaining traction astatine singular speed. Since Fabric launched 2 years ago, it has grown faster than immoderate different information and analytics level successful the industry. More than 28,000 customers—including 80% of the Fortune 500—now trust connected Fabric, and its ecosystem continues to grow arsenic partners physique solutions to lick the astir analyzable information challenges.

Another leap guardant crossed Microsoft Databases and Microsoft Fabric

As Fabric becomes the central connection point for data, we’re strengthening the database layer at the heart of your information estate—ensuring you person the standard and performance required for AI.  

Microsoft already offers 1 of the industry’s astir broad database portfolios, and we’re expanding it adjacent further—while profoundly integrating these capabilities into Fabric. I’m excited to denote the wide availability of SQL Server 2025, Azure DocumentDB, and SQL database and Cosmos DB successful Fabric, on with the preview of our newest addition, Azure HorizonDB. With these new offerings, you have a world-class database option to build erstwhile and deploy astatine the edge, arsenic level arsenic a work (PaaS), or adjacent arsenic bundle arsenic a work (SaaS). And due to the fact that our full portfolio is either Fabric-connected oregon Fabric native, Fabric serves arsenic a unified hub for your full information estate. Below I’ll cover how these caller databases are purpose-built to enactment your AI projects.  

Deploy the adjacent procreation of Microsoft Databases

Modernize your SQL property with SQL Server 2025, present mostly available

Microsoft has been shaping the SQL scenery for much than 35 years. Now, with the merchandise of SQL Server 2025 into wide availability, we’re introducing the adjacent evolution—one that brings developer‑first AI capabilities astatine the edge, wrong the acquainted T‑SQL experience. Smarter hunt combines precocious semantic quality with full‑text filtering to uncover richer insights from analyzable data. AI exemplary absorption utilizing exemplary definitions successful T-SQL allows seamless integration with fashionable AI services specified arsenic Microsoft Foundry.

Enterprise reliability and information stay best-in-class. Enhanced query performance, optimized locking, and improved failover assistance guarantee higher concurrency and uptime for mission‑critical workloads. With strengthened credential absorption done Microsoft Entra ID via Azure Arc, SQL Server 2025 is unafraid by design. Your information is besides instantly accessible for your AI and analytics successful Microsoft OneLake with mirroring for SQL Server 2025 successful Fabric, present besides mostly available.

SQL Server 2025 is the astir important merchandise for SQL developers successful a decade. And the effect to our preview has been overwhelming, with 10,000 organizations participating, 100,000 databases already deployed, and download complaint 2 times higher than SQL Server 2022. If you privation to articulation each those who’ve already adopted SQL Server 2025, download it today.

Azure DocumentDB: MongoDB-compatible, AI-ready, and built for hybrid and multi-cloud

We’re excited to denote Azure DocumentDB, a caller work built connected the open-source, MongoDB-compatible DocumentDB modular governed by the Linux Foundation. The archetypal Azure managed work to enactment multi-cloud and hybrid NoSQL, Azure DocumentDB tin tally consistently crossed Azure, on-premises, and different clouds.

Azure DocumentDB gives you the state to clasp unfastened root portion achieving scale, security, and simplicity. It’s AI-ready, with capabilities similar vector and hybrid hunt to present much applicable results. Instant autoscale meets demand, and autarkic compute and retention scaling keeps workloads efficient. Security and availability is standard, with Microsoft Entra ID integration, customer-managed encryption keys, 35-day backups included, and a 99.995% availability service-level statement (SLA). And soon, enhanced full-text hunt volition adhd features similar fuzzy matching, proximity queries, and expanded connection support, making it adjacent easier to physique intelligent, search-driven apps.

Azure DocumentDB is present mostly available, truthful you tin try it today. You tin besides larn much astir Azure DocumentDB and each the Azure Database quality by speechmaking Shireesh Thota’s, Corporate Vice President of Azure Databases, announcement blog.

Azure HorizonDB: PostgreSQL designed for your mission-critical workloads

PostgreSQL has go the backbone of modern information solutions acknowledgment to its affluent ecosystem, extensibility, and unfastened root foundation. Microsoft is arrogant to beryllium the #1 PostgreSQL committer among hyperscalers, and we’re gathering connected that enactment with Azure HorizonDB.

Now successful aboriginal preview, Azure HorizonDB is simply a afloat managed, PostgreSQL-compatible database service, built to grip the standard and show required by the modern enterprise. It goes acold beyond unfastened root Postgres, with auto-scaling retention up to 128 TB, scale-out compute up to 3,072 vCores, <1 millisecond multi-zone perpetrate latency, and endeavor information and compliance. Vector hunt is built-in, on with integrated AI exemplary absorption and seamless connectivity to Microsoft Foundry truthful you tin physique modern AI apps. Combined with GitHub Copilot, Fabric, and Visual Studio Code integrations, it provides an intelligent and unafraid instauration for gathering and modernizing applications astatine immoderate scale. To larn much astir Azure HorizonDB, work our announcement blog.

Accelerate app improvement with Fabric SaaS Databases, present mostly available

We are besides releasing a caller people of SaaS databases, some SQL database and Cosmos DB successful Fabric, into wide availability. Data developers present person entree to world-class database engines wrong the aforesaid unified level that powers analytics, AI, and concern intelligence.

Fabric Databases are designed to streamline your exertion development. You tin proviso them successful seconds, and they don’t necessitate the accustomed granular configuration oregon heavy database expertise. They supply enterprise-grade performance, are unafraid by default with features similar unreality authentication, customer-managed keys, and database encryption, and travel natively integrated into the Fabric platform, adjacent utilizing the aforesaid Fabric capableness units for billing.

With Fabric databases, developers present person the flexibility to physique applications grounded successful operational, transactional, and analytical data. Together, these offerings marque Fabric a developer-first information level that is streamlined, scalable, and acceptable for modern information applications.

Learn much by speechmaking Shireesh Thota’s, Corporate Vice President of Azure Databases, announcement blog.

All your databases connected into Fabric

We’re making it easier than ever to enactment with your full Microsoft database portfolio successful Fabric, giving you a single, unified spot to negociate and usage each your data. Building connected our existing mirroring enactment for Azure SQL Database and Azure SQL MI, we’re present announcing the wide availability of mirroring for Azure Database for PostgreSQL, Azure Cosmos DB, and SQL Server versions 2016–2022 and 2025. With these databases mirrored straight into Fabric, you tin destruct accepted extract, transform, and load (ETL) pipelines and marque your information instantly acceptable for analytics and AI.

Getting your information property acceptable for AI with Microsoft Fabric

Choosing the close database is essential, but it’s conscionable the beginning. The large accidental lies successful driving frontier transformation, wherever information becomes the instauration for an AI-native enterprise. We urge focusing connected 3 halfway steps:

  • Unifying your information estate to destruct silos and complexity.
  • Creating semantic meaning truthful your information is acceptable for AI.
  • Empowering agents to enactment connected insights and alteration operations.

In this section, I’ll dive into the latest enhancements to Microsoft Fabric that assistance you execute each measurement of your information journey. This includes expanded interoperability successful OneLake with SAP, Salesforce, Azure Databricks, and Snowflake, the instauration of Fabric IQ—a caller workload that adds semantic understanding—and enhanced agentic capabilities crossed Fabric to assistance you physique richer, AI-powered information experiences.

This is the aboriginal of data, and it’s already wrong reach. With Fabric and our database innovations, Microsoft is helping organizations determination seamlessly from penetration to action—unlocking the afloat imaginable of your information and the AI built connected apical of it.

Unify your information property with Microsoft OneLake

Microsoft OneLake unifies each your data—across clouds, on-premises, and beyond Microsoft—into a azygous information water with zero-ETL capabilities similar shortcuts and mirroring. Alongside the further mirroring sources for Microsoft Databases, we’re besides introducing the preview of shortcuts to SharePoint and OneDrive. This allows you to bring unstructured productivity information into OneLake without copying files oregon gathering ETL pipelines, making it easier to bid agents and enrich your structured data.

Once connected to OneLake, your information becomes easy discoverable successful the apps your teams usage each time similar Power BI, Teams, Excel, Copilot Studio, and Microsoft Foundry. Today, we are taking that a measurement further with autochthonal integration with Foundry IQ—the adjacent procreation of retrieval-augmented procreation (RAG). Agents trust connected context—Foundry IQ’s cognition bases present high-value discourse to agents by simplifying entree to aggregate information sources and making connections crossed information. You tin usage the OneLake cognition root successful Foundry IQ to link agents to multi-cloud sources similar AWS S3, on-premises sources, and structured and unstructured data.

See however shortcuts and mirroring unify your information successful OneLake and substance the adjacent procreation of intelligent agents successful Microsoft Foundry:

Expanding OneLake interoperability with starring information platforms

We are besides seeing large momentum with dozens of partners extracurricular of Microsoft profoundly integrating with OneLake, including ClickHouse, Dremio, Confluent, EON, and galore more. And now, we are thrilled to adhd new, deeper interoperability with SAP, Salesforce, Azure Databricks, and Snowflake.

First, we’re deepening interoperability with the systems organizations trust connected most, SAP and Salesforce. With the motorboat of SAP Business Data Cloud Connect for Microsoft Fabric, customers tin let bidirectional, zero-copy information sharing betwixt SAP Business Data Cloud (BDC) and Fabric. At the aforesaid time, we are moving with Salesforce to integrate their information into Fabric utilizing the aforesaid zero-copy approach, unlocking precocious analytics and AI capabilities without the overhead of accepted ETL.

We’re besides strengthening interoperability with Azure Databricks and Snowflake truthful you tin usage a azygous transcript of information crossed platforms. By the extremity of 2025, Azure Databricks volition release, successful preview, the quality to natively work information from OneLake done Unity Catalog, enabling seamless entree without duplication oregon analyzable information movement. Looking ahead, Databricks volition besides adhd enactment for penning to and storing information straight successful OneLake, allowing afloat two-way interoperability. Read much astir this interoperability.

Our collaboration with Snowflake connected bidirectional information entree continues arsenic well. We are introducing a caller point successful OneLake called a Snowflake Database and a caller UI successful Snowflake—both designed to let OneLake to beryllium the autochthonal retention solution for your Snowflake data. We’re besides bringing Snowflake mirroring to wide availability, allowing you to virtualize your outer Snowflake-managed Iceberg tables successful OneLake with shortcuts created and handled automatically. Together, these innovations fto you tally immoderate Fabric workload—whether analytics, AI, oregon visualization—directly connected your Snowflake-managed Iceberg tables.

Learn much astir our Snowflake collaboration by speechmaking our latest joint blog oregon by watching the pursuing demo:

Finally, in adjacent collaboration with dbt Labs, we are besides excited to denote built-in enactment for their manufacture starring information translation capability. Now successful preview, dbt jobs successful Microsoft Fabric fto you build, test, and orchestrate dbt workflows successful your Fabric workspaces. Learn much successful this blog.

Create semantic cognition to substance AI with Fabric IQ

As Frontier Firms bid agents connected their endeavor data, it’s go wide that prime and discourse substance much than information volume. Agents request concern discourse crossed relationships, hierarchies, and meaning to crook earthy information into actionable insight. That’s wherefore we’re introducing Fabric IQ—a caller workload designed to representation your datasets to the real-world entities they represent, creating a shared semantic operation connected apical of your data.

The powerfulness of IQ lies successful however it unifies disparate information types nether a single, coherent framework. Built upon Power BI’s industry-leading, affluent semantic exemplary technology, IQ brings unneurotic analytical data, time-series telemetry, and geospatial information, each organized nether a semantic model of concern entities and their relationships, properties, rules, and actions. You tin past make operations agents, a caller benignant of cause successful Fabric, which tin usage this exemplary to enactment arsenic virtual squad members, monitoring real-time information sources, identifying patterns, and taking proactive action. Instead of forcing your teams and adjacent agents to deliberation successful presumption of tables and schemas, IQ allows you to align information with however your enactment operates.

In short, Fabric IQ is designed to exemplary world with data, truthful that each insight, prediction, and enactment is grounded successful however your enactment really operates. You tin larn much astir IQ successful Yitzhak Kesselman’s, Corporate Vice President of Messaging and Real-Time Intelligence, announcement blog.

Empower data-rich agents with Copilot, Fabric information agents, and operations agents

As organizations standard their AI initiatives, the quality to link intelligent agents with enterprise-grade information is becoming a captious differentiator. Fabric is making this imaginable with a acceptable of integrated AI experiences: Copilot successful Power BI helps you inquire questions of your data, Fabric information agents let deeper analysis, and the caller Fabric operations agents fto you show your information property and instrumentality enactment successful existent time. These experiences tin beryllium utilized crossed Fabric oregon arsenic foundational cognition sources successful industry-leading AI tools similar Microsoft Foundry, Copilot Studio oregon adjacent Microsoft 365 Copilot to powerfulness smarter, much data-rich AI experiences.

Beyond introducing operations agents arsenic portion of Fabric IQ, we’re besides expanding what information agents and Copilot tin do. Along with existing integration with Microsoft Foundry and Copilot Studio, Fabric information agents tin present beryllium embedded straight successful Microsoft 365 Copilot. This lets concern users (with the close permissions) entree trusted cognition from OneLake and transforms Microsoft 365 from a productivity suite into an intelligent insights platform.

They tin besides enactment arsenic hosted Model Context Protocol (MCP) servers, making it casual to integrate with different applications and agents crossed the AI ecosystem. Finally, information agents tin present crushed crossed some structured and unstructured data. Thanks to an integration with Azure AI Search, information teams tin adhd their existing unstructured information hunt endpoints arsenic a root successful information agents. Learn much the Fabric information cause enhancements by speechmaking the Fabric AI blog.

We’re besides enhancing the standalone acquisition for Copilot successful Power BI with a caller hunt experience. Simply picture what you need, and Copilot volition find the applicable report, semantic model, oregon information cause and aboveground the close answers. This standalone acquisition is besides coming to Power BI mobile truthful you tin usage it connected the go.

Take a look astatine however you tin use each of these AI experiences unneurotic seamlessly:

In short, we’re redefining what it means to person an AI-powered information estate. With information agents, Copilot successful Power BI, and operations agents successful Fabric IQ, AI is present woven crossed Fabric. And with autochthonal integration to Microsoft Foundry and Copilot Studio, you tin easy adhd Fabric agents arsenic gathering blocks to make much intelligent, informed customized agents.

You besides tin spot much innovation coming to the Fabric level by speechmaking Kim Manis’, Corporate Vice President of the Fabric Platform, Fabric blog oregon by checking retired the much method Fabric November 2025 Feature summary blog.

Mark your calendar for FabCon and SQLCon

We are excited to denote SQLCon 2026, which volition hap astatine the aforesaid clip and the aforesaid determination arsenic the Microsoft Fabric Community Conference (FabCon), happening March 16–20, 2026 successful Atlanta, Georgia. By uniting the powerhouse SQL and Fabric communities, we’re giving information professionals everyplace a unsocial accidental to maestro the latest innovations, stock applicable knowledge, and accelerate what’s imaginable with information and AI, each successful 1 almighty week. Register for either league and bask afloat entree to both, with the flexibility to premix and lucifer sessions, keynotes, and assemblage events to acceptable your interests.

Watch these announcements successful enactment astatine Microsoft Ignite

If you’re funny successful seeing these announcements live, I promote you to articulation my Ignite session, “Innovation Session: Microsoft Fabric and Azure Databases – the information property for AI” either successful idiosyncratic oregon online astatine nary cost. I’ll not lone screen these large announcements but amusement you however they travel unneurotic to assistance you make a unified, intelligent information instauration for AI.

You tin besides dive deeper into these announcements and truthful overmuch much by watching the remainder of the breakout sessions crossed Azure Data:

Tuesday, November 18

Wednesday, November 19

Thursday, November 20

Read Entire Article