Unleashing the Future of AI with Microsoft Fabric and Azure Data

In this article, I’ll walk you through the newest updates to Microsoft Fabric and the wider Azure Data portfolio. These updates are crafted to help every developer make the most out of their data with more flexibility and power than ever before.

Microsoft’s latest annual Work Trend Index highlighted a fresh style of organisation, completely redesigned by artificial intelligence. These so-called frontier firms are shaping the future with AI agents, mixing teams of humans and digital assistants, and even building whole agent-driven teams steered by people. As companies step into this new age, these agents will span departments and roles, handling routine work and delivering timely, personalised insights organisation-wide.

Developers are the driving force behind this wave of AI-powered agents. To make these agents function effectively, developers need much more than AI models; they must harness all kinds of business data—analytics, transactions, operations—no matter the format or structure. For example, The Estée Lauder Companies used a digital agent trained on structured consumer insights to provide staff with actionable information instantly. Meanwhile, dentsu relies on a group of AI agents educated with unstructured HR and compliance data, helping answer employees’ questions without delay.

Having the best tools is crucial for developers aiming to unify and prep data for artificial intelligence and digital agents. Microsoft ensures you’re fully equipped, whether you’re working with analytics, operational records, or transactional data. Microsoft stands as a leader in the 2024 Gartner® Magic Quadrant for Data Integration Tools, Cloud Database Management Systems, Cloud AI Developer Services, as well as Analytics and Business Intelligence Platforms.

In this article, you’ll discover the most recent announcements for Microsoft Fabric and the Azure Data platform, each designed to help you achieve more as a developer with your data:

Fabric announcements: Cosmos DB (NoSQL) is being added to Fabric, along with a digital twin builder and new ways to use Copilot to interact directly with your data.

Azure Data announcements: News about SQL Server 2025, PostgreSQL support in VS Code with GitHub Copilot, AI Foundry integration with Azure Cosmos DB and Azure Databricks, and SAP links with Azure Databricks.

Microsoft Fabric combines every feature and tool you need under one AI-powered platform, helping your teams tackle any data challenge. Our mission is to combine your essential data services into a single system—open, flexible, and fully integrated—eliminating the need for patchwork solutions. This vision already resonates with over 21,000 customers, including over 70% of Fortune 500 firms, who now rely on Fabric. In fact, more than half of our users take advantage of three or more workloads, finding real success with this all-in-one data platform.

By bringing together Microsoft’s best-in-class tools inside an easy-to-use software-as-a-service (SaaS) platform, developers can move their projects faster. Whether you’re managing a database, streaming real-time data, or building machine learning models, Fabric offers seamless, immediate access. With SaaS, you skip infrastructure headaches and can start building immediately. Integrated CI/CD features and direct connections to everyday utilities like GitHub and VS Code make development smoother for every developer.

Importantly, Fabric lets you unlock any type of data for analytics and beyond. With Microsoft OneLake—Fabric’s universal data lake—you can access structured or unstructured data across any cloud or format. Fabric Real-Time Intelligence streamlines data in constant motion. The launch of Fabric Databases now covers your transactional workloads, too. And now, with Cosmos DB in Fabric, operational and semi-structured data support is further enhanced.

Accelerate app development with Cosmos DB in Fabric

Just six months ago, we introduced Fabric Databases—a new breed of SaaS databases fully embedded within Fabric. These databases are simple to deploy, easy to manage, and instantly accessible, giving developers a simple way to streamline building new apps. We started with SQL databases, which provide quick, robust, and secure solutions for structured data.

Agents take on more sophisticated jobs quickly, so handling semi-structured content like text, documents, and emails becomes increasingly important for AI training. That’s why we’re excited to extend Fabric Databases to support semi-structured information with a preview of Cosmos DB in Fabric. A Bloomberg CIO survey even named Azure Cosmos DB as the top pick for building generative AI apps.¹ In fact,

If you run into issues connecting Cosmos DB to Fabric, check your access permissions first. Make sure your account has the right roles assigned. Next, double-check that your databases and resources are in the correct region and network for seamless integration. If you’re struggling with data sync or format mismatches, refer to the Microsoft Fabric documentation for correct setup procedures. This approach ensures Cosmos DB can unlock the full potential of your data while keeping connections smooth and secure.

Azure Cosmos DB in Fabric powers ChatGPT’s massive scale, handling trillions of transactions and storing petabytes of data for over 500 million weekly users.²

Now, this high-performance database technology is coming to Microsoft Fabric. With Azure Cosmos DB in Fabric, developers can launch a robust database in just a few steps. Enjoy built-in, enterprise-level dynamic scalability and remarkable 99.999% reliability. Supporting both SQL and NoSQL options, Azure Cosmos DB in Fabric gives you flexible tools to design AI solutions rooted in operational, transactional, or analytical data. Best of all, your Cosmos DB data syncs instantly to OneLake—perfect for running near real-time analytics, like sentiment detection for chat platforms. Below, see these powerful features demonstrated:

Get started with Azure Cosmos DB in Fabric today, or discover more by reading the Cosmos DB in Fabric blog.

Bring the Physical World to Life with Digital Twin Builder in Fabric

Once your data sits in Microsoft Fabric, you gain richer tools for analysis. One standout is the new digital twin builder, now available for preview. This innovative feature lets businesses design, control, and visualise digital counterparts of both real and logical objects at scale. Digital twin builder lives within Fabric Real-Time Intelligence and offers a simple, low-code (or even no-code) way to manage these digital models.

With digital twin builder, connect and orchestrate virtual representations of anything—from physical equipment and customer records to entire manufacturing processes. Use these models to drive advanced analytics, simulate what-if scenarios, or automate tasks. To dive deeper, visit the Digital Twin Builder in Fabric blog or start your free 60-day Microsoft Fabric trial. Watch the demo below to see digital twin builder in action.

Let Everyone Explore Their Data with Power BI & Microsoft Copilot Studio

Data only makes an impact when it empowers people to take action. Making information accessible—especially through natural chat experiences—has rapidly become a key AI advancement. With a smart agent, teams can easily explore their datasets, driving better decisions through simple questions and instant answers.

Chat with Your Data Using Copilot in Power BI

A new, full-screen Copilot view in Power BI now lets anyone engage with the data they’re permitted to see. Over 30 million rely on Power BI every month, and it’s tightly woven into daily business tools like Microsoft 365. Soon, users can open Copilot right from their home screen, pose broad questions, and receive accurate answers. Copilot scans across your accessible reports and data models, delivering on-point responses. Now, you can even use Copilot in Power BI directly within Teams to keep productivity high. See it in action here:

Power BI is also getting deeper integration with Microsoft 365 Copilot. Soon, the Power BI agent will help you find content, ask questions, and explore visuals, all within Microsoft 365 Copilot—no context switching required.

Link Fabric Data Agents into Microsoft Copilot Studio

Within Fabric, you can design data agents—virtual analysts who field your questions in natural language. Excitingly, these agents now integrate with Copilot Studio. Deploy them easily across Microsoft Teams or Microsoft 365 Copilot, and let them interpret complex data or surface insights instantly—all while respecting your access permissions. They’ll even automate routine jobs, such as sending updates or starting workflows, which streamlines how teams tap into Azure Cosmos DB in Fabric and make decisions.

Troubleshooting Guide: Azure Cosmos DB in Fabric

If you’re having trouble getting started with Azure Cosmos DB in Fabric, consider these steps:

Deployment Issues: Check your required permissions and that your Fabric workspace supports Cosmos DB.
Connection Errors: Ensure the right connection string and access keys are used. Firewall rules may need updating.
Real-Time Sync Delays: Confirm data ingestion settings and validate that Cosmos DB integrates properly with OneLake for real-time results.
NoSQL/SQL Query Problems: Review syntax and confirm model compatibility—switch between SQL and NoSQL as needed using guides from the Azure Cosmos DB in Fabric documentation.

Still facing problems? Review the official Cosmos DB in Fabric blog for advanced fixes and updates.

Explore the latest chat with your data experience in Power BI. Plus, keep an eye out for the soon-to-arrive integration of Fabric data agents with Copilot Studio. Exciting capabilities are on the horizon!

Gain a Competitive Advantage with Azure Data

Reliable databases have long been the foundation for enterprise information management—storing, retrieving, and analysing the data powering modern business. As we move forward into an AI-driven future, Azure Data becomes even more vital for enabling intelligent, adaptive applications.

The ODP Corporation—including Office Depot—demonstrates Azure Data’s competitive advantage. They utilise Azure Cosmos DB paired with Azure AI services to provide HR information instantly to staff, slashing turnaround from 24 hours to real time.

“We rely on Azure Cosmos DB for storing HR data at the profile level. It gives us a real-time link between the back end and front end—every user receives a profile which the model updates live, ensuring context is never lost.”

—Mick Feller, Distinguished Engineer, The ODP Corporation

ODP’s experience only scratches the surface. Organisations such as DocuSign, BNY Mellon, and Mondra also drive their own AI revolutions by leveraging Azure Data solutions for a competitive advantage.

We are continually enhancing our Azure Data services to keep up with AI innovation and support your evolving business applications. Today, explore significant improvements across our advanced SQL and open-source database options, all built to help developers create intelligent solutions easily.

Secure AI Apps from On-Premises to Cloud with SQL Server 2025

SQL Server has been trusted for over 35 years for secure and high-performance data management. Now, SQL Server 2025 enters public preview, purpose-built for the age of AI applications.

The new SQL Server 2025 transforms into a robust vector database, offering native support for secure and advanced AI features. You’ll find enhanced data filtering, built-in vector search, and seamless integration with popular AI frameworks like LangChain and Semantic Kernel. Whether using local or cloud-based models, SQL Server 2025 streamlines workflows by supporting direct vector embedding and compatibility with Entity Framework Core.

Our early adopters speak volumes: over 3,400 applicants joined the private preview, and adoption is happening twice as fast as with SQL Server 2022. Get started seamlessly with integration to Microsoft Fabric and Azure Arc, plus experience real-time data mirroring in Fabric. To learn more, see SQL Server 2025 in action and discover how to get involved today.

Azure PostgreSQL, VS Code & GitHub Copilot: An Integrated Approach

Open-source databases shape the future of intelligent apps, with PostgreSQL leading the way, as highlighted in Stack Overflow’s 2024 Developer Survey. Microsoft actively supports this open-source community, boasting the highest number of Postgres contributors among major cloud providers.

We’re excited to preview the enhanced PostgreSQL extension for VS Code, integrated with GitHub Copilot. This powerful tool simplifies database management in VS Code—whether developing with Azure, Docker, or on-premises. The Copilot integration brings AI-powered guidance to your fingertips, making it easy to design schemas, optimise SQL, or resolve issues directly in your editor.

Additionally, DiskANN is now generally available in Azure Database for PostgreSQL. DiskANN offers one of the fastest vector indexing systems available, allowing you to build scalable, low-latency AI apps that exceed the speed of pgvector indexing.

While vector search is crucial for generative AI, it may struggle with interpreting some complex semantic relationships within business data. To expand these capabilities, we’re introducing new AI semantic operators in Azure PostgreSQL, now in preview. These operators use large language models directly inside the database, revealing richer insights and deeper context from your operational data.

Direct Integration: Azure Cosmos DB and Databricks with Azure AI Foundry

Developers aiming to design bespoke AI apps will be pleased to know you can now connect Azure Cosmos DB directly to Azure AI Foundry for powering AI-driven solutions.

With this capability, business users can securely store and manage conversations between users and AI agents within Cosmos DB—using the dedicated Azure AI Foundry SDK. This enables agents to recall past exchanges, maintaining continuity and providing a seamless conversational experience. Thread storage features are now generally available for everyone.

Soon, you’ll also be able to harness the data within your Cosmos DB accounts for AI solutions on Azure AI Foundry. This opens up real-time access to operational data, positioning Cosmos DB as the first Azure database capable of powering agents and intelligence models within Azure AI Foundry.

Troubleshooting Tips for Azure Data

If you face integration or performance issues, try the following steps:

Check connections: Ensure your Azure Data services are correctly linked across environments.
Monitor usage: Use Azure Monitor to spot performance bottlenecks or timeouts in your databases.
Review logs: Error logs in SQL Server and PostgreSQL extensions often provide clear hints about setup or connection problems.
Update extensions: Make sure you’re using the latest VS Code extension versions for optimal Azure Data integration.
Consult Azure docs: Look up Microsoft’s troubleshooting guides if custom setups are failing to connect or perform.

For persistent problems, consider isolating new features or rolling back to a known working configuration. This helps identify conflicts caused by recent changes.

We are excited to introduce the public preview of the Azure AI Foundry connection for Azure Databricks. This new link focuses on two vital scenarios: enabling Foundry Agents to harness AI/BI Genie and executing Azure Databricks Jobs. Together, these features drive smarter knowledge search and allow Foundry Agents to provide answers rooted in your business data. As much of the world’s information lives within organisations, this development is key for efficient enterprise solutions. For those eager to explore all the recent Azure Databricks updates, take a look at the detailed Azure blog.

Expanding Our Databricks and SAP Partnership on Azure Databricks

We’re delighted to announce an even stronger alliance between SAP and Databricks as SAP launches Business Data Cloud (BDC) and SAP Databricks on Azure, coming in Q3 2025. SAP Databricks will underpin SAP Business Data Cloud, empowering businesses to access Databricks analytics tools, finely tuned for SAP workloads. Now, your development teams can innovate using the core business data managed in SAP. Thanks to first-party Microsoft Azure Databricks services, this integration promises better performance, scalability, and security, tightly woven into native Azure Databricks capabilities.

Connect Your Azure Databricks Data to Fabric Easily

Mirroring now lets you reach your Azure Databricks Unity Catalog tables via OneLake—this feature is available in preview. You just add your Databricks workspace URL, pick your desired catalog, and Fabric takes care of creating shortcuts for every table. Changes stay almost instantly in sync, helping you work with up-to-date data across platforms. Dive deeper by watching the step-by-step video or explore further with the official documentation.

How to Troubleshoot Azure Databricks and Fabric Connections

If you encounter issues with syncing your Azure Databricks data to Fabric, try these steps:

Double-check your Unity Catalogue permissions and make sure your workspace URL is correct.
If tables do not appear, confirm that mirroring is enabled and your chosen catalogue contains supported data formats.
For syncing issues, refresh connections in Fabric and inspect the event logs for any errors.
Should you keep experiencing trouble, refer to Fabric’s troubleshooting guide for Azure Databricks.

A proactive check of credentials and connection settings usually solves most problems quickly.

Moving Forward with Data and AI Innovations in Azure Databricks

These new releases mark a bold step forward for data and AI innovation within Azure Databricks. Our goal is to enable seamless integration between your enterprise data and advanced artificial intelligence applications. No matter where your information sits, or what kind of data you store, Azure Databricks ensures your team can break down silos and drive smarter decisions.

Watch Azure Databricks Announcements LIVE at Microsoft Build

Join us at Microsoft Build from 19 to 22 May 2025 to witness Azure Databricks in action! Here’s a taste of the featured sessions:

BRK206: Microsoft Fabric for Developers: Build scalable data and AI solutions.
BRK204: What’s new in Microsoft Databases: Empowering AI-Driven App Development.
BRK202: Scale and secure MongoDB-compatible apps with Azure Cosmos DB.
BRK203: Get faster LLM responses and lower latency with Azure Managed Redis.
BRK205: What’s coming in Fabric Automation and CI/CD.
BRK207: SQL Server 2025: The Database Developer Reimagined.
BRK208: Building AI agents for actionable insights using data in Fabric.
BRK209: Building real solutions with Real-Time Intelligence in Fabric.
BRK210: Build AI apps and unlock the power of your data with Azure Databricks.
BRK211: Building advanced agentic apps with PostgreSQL on Azure.
BRK212: Design scalable data layers for multi-tenant apps with Azure Cosmos DB.
BRK213: Enable advanced AI scenarios with Unified Data Estates in Microsoft Fabric.

References

¹ Bloomberg, 60% of Enterprise Chief Information Officers Report Plans to Increase Spending on AI Inferencing Workloads with Microsoft, according to a Bloomberg Intelligence Survey

² Forbes, ChatGPT Hits 1 Billion Users? ‘Doubled In Just Weeks’ Says OpenAI CEO

³ Stack Overflow, 2024 Developer Survey