Held in Las Vegas, FabCon 2025 brought together the global Microsoft Fabric community from around the world for an exciting week of announcements, technical deep dives, and product showcases. This year’s conference marked a major milestone for the platform, introducing what many attendees, including Microsoft’s own product leaders, called the most significant wave of updates since Fabric’s General Availability.
The announcements touched nearly every aspect of the platform, with a strong focus on four key areas: AI, security, interoperability, and productivity. From bringing Copilot to every paid SKU, to introducing centralized OneLake Security, and expanding CI/CD support across services like Eventstream and Data Warehousing, the microsoft fabric community conference showcased just how fast Fabric is evolving to meet the needs of modern data teams.
In this article, we’ll break down the highlights from Microsoft Fabric Community Conference 2025, focusing on the updates that matter most for technical teams working with Microsoft Fabric.
FabCon 2025 delivered breakthrough after breakthrough, showcasing the latest features and a multitude of impactful updates unveiled. Azure AI was prominently featured, highlighting its integration within Microsoft's suite of data and AI products. In this section, our data experts recommend focusing on two standout announcements: OneLake Security Updates and Copilot & AI Everywhere.
One of the headline updates at Microsoft Fabric Community Conference was the rollout of OneLake Security, a long-anticipated capability that fundamentally redefines how access control is managed across Microsoft Fabric. This update is key for business intelligence, data management, and analytics, as it introduces a centralized, unified security model designed to reduce complexity, ensure consistency, and elevate data governance across the platform.
Historically, managing security in Fabric meant juggling separate configurations in workspaces, semantic models, warehouses, and OneLake folders. With OneLake Security, all access permissions can now be defined and managed in a single, unified interface—dramatically simplifying security administration across the platform.
A new OneLake Security UI was also announced at this year’s Microsoft Fabric Community Conference (FabCon 2025), making it easier to create roles, assign members, and configure data access for each role. Users can also close modal windows for managing security configurations.
Access policies defined in OneLake Security are automatically enforced across Spark, SQL Endpoints, and Power BI in Direct Lake mode. Newly created items use OneLake Security by default, and existing items can be migrated to adopt it. This eliminates the need for separate rules in each engine and ensures that access policies behave the same way everywhere.
For example:
Another interesting announcement at the Microsoft Fabric Community Conference is that fine-grained security controls are now embedded directly into OneLake, allowing you to define row-level and column-level access for each role. This means sensitive data can be secured without needing to manage complex permissions within each analytical engine.
With the new UI, configuring this level of access is straightforward:
Security rules defined in OneLake are automatically inherited when users access data through tools like Power BI reports, notebooks, Excel, or SQL clients. This inheritance model reduces complexity and ensures consistent user experience across different environments.
As announced at Microsoft Fabric Community Conference in Las Vegas, OneLake Security is in limited preview, with early access available via sign-up. Once enabled for a workspace, new capabilities—including the updated security UI and support for row/column-level controls—will become available for testing.
OneLake Security marks a major shift in how access is managed in Microsoft Fabric. It delivers on long-requested features while laying the groundwork for deeper data integration and improvements ahead.
Key benefits include:
At the Microsoft Fabric Community Conference they also shared that additional enhancements are on the roadmap, including deeper integration, performance improvements, and broader support across more workloads.
At Microsoft Fabric Community Conference 2025, Microsoft made it clear: AI products in Fabric are no longer limited to enterprise-scale deployments. The product team announced a sweeping update that brings Copilot and AI features to all paid Microsoft Fabric SKUs, including the lower-tier capacities starting from F2 and up — a major step toward democratizing AI within the platform. Attendees were eager to learn about the new AI capabilities.
Previously available only in higher-capacity SKUs like F64, Copilot is now rolling out to every Microsoft Fabric capacity level. This change, expected to be completed by end of April 2025, according to information from the Microsoft Fabric Community Conference, ensures that teams of all sizes — not just large enterprises — can access Copilot’s AI-powered capabilities for tasks like:
This update announced at Las Vegas dramatically expands the reach of generative AI within Microsoft Fabric, enabling broader organizational use and supporting more users in becoming data driven.
Another key announcement from Microsoft Fabric Community Conference was the expansion and rebranding of Fabric’s AI Skills to Fabric Data Agents. These conversational, AI-powered assistants help users interact with data in natural language, offering a more intuitive way to explore and work with datasets — without requiring deep technical expertise.
Now available for more data sources, including:
Data agents can now be embedded in applications via published endpoints, opening new possibilities for enterprise and citizen developers alike.
At the Microsoft Fabric Community Conference, they also showcased the out-of-the-box AI functions available directly within Microsoft Fabric:
These functions allow users to integrate natural language processing tasks into data pipelines, notebooks, or Spark transformations with minimal setup — helping analysts and engineers apply AI logic faster and more efficiently.
FabCon 2025 introduced a Copilot experience inside Microsoft Fabric Notebooks, where it assists users with writing code, transforming data, and performing analysis directly within the notebook interface. This improves the developer workflow by making large language models easily accessible during coding sessions.
In addition, AI functions for Spark are now in preview, allowing teams to apply LLM-powered transformations to data at scale — a major productivity boost for teams working on large datasets.
Another powerful new capability announced at Microsoft Fabric Community Conference is Fabric User Data Functions. These functions let developers and data engineers:
User data functions help teams standardize data engineering workflows and reduce repetitive effort by turning custom routines into reusable, centrally managed assets.
Microsoft Fabric now includes native connectivity to Azure OpenAI Service, giving users direct access to cutting-edge large language models within their Fabric environment. This integration opens up new opportunities to embed LLMs into data pipelines, transformations, and user-facing applications.
Together, these updates announced at Microsoft Fabric Community Conference represent a significant leap forward in Fabric’s AI capabilities — not just in power, but in accessibility. With Copilot and AI tools becoming standard across all tiers, Microsoft is pushing Fabric toward becoming a truly AI-native analytics platform for every team, not just the few.
The Microsoft Fabric Community Conference made it clear: Microsoft is making it easier than ever to adopt Fabric databases. From a streamlined Migration Assistant for Azure Synapse users to powerful new Mirroring features, the focus is on reducing complexity and accelerating time to value. Here are two standout updates that help teams move faster and operate more efficiently.
Migrating from Synapse to Fabric just became a whole lot easier. At the Microsoft Fabric Community Conference 2025, Microsoft introduced the preview of a native Migration Assistant, designed specifically to support Azure Synapse Analytics Dedicated SQL Pool customers in connecting and transitioning seamlessly to Microsoft Fabric.
This experience is now being rolled out globally and is integrated directly into the Microsoft Fabric interface, simplifying what was once a complex, multi-step process. The migration flow is broken into four guided phases:
Using a DACPAC file (a snapshot of your database schema), the assistant helps migrate database objects like tables, views, stored procedures, functions, and security roles. T-SQL code is automatically translated into Microsoft Fabric-compatible syntax, with adjustments logged and shown for review.
If any metadata fails to migrate (due to unsupported syntax or compatibility issues), the assistant helps identify errors, dependencies, and adjustments. You can fix these scripts manually or use Copilot to suggest and apply corrections — complete with comments explaining what was changed and why.
After metadata is successfully migrated, the built-in copy job wizard lets users transfer actual data from the source system. Users can configure:
In the final step, users are prompted to reconnect ETL tools and reporting systems to the new Microsoft Fabric warehouse. Once completed, the new warehouse is fully operational and ready for production use.
This end-to-end workflow allows organizations to migrate both their database structure and data in a structured, auditable, and assisted way — reducing risk, cost, and time to value.
At the Microsoft Fabric Community Conference they also shared upcoming improvements planned for the Migration Assistant, including:
These updates are paving the way forward suggesting that Microsoft has a strong commitment to making Fabric de natural next step for existing Synapse customers.
If you were impressed by the migration features presented at Microsoft fabric community conference and you'd like to migrate your data from Azure Synapse to Microsoft Fabric Data Warehouse, you can consider partnering up with us as your Microsoft Fabric consulting experts.
Mirroring in Microsoft Fabric took a significant step forward at Microsoft Fabric Community Conference (FabCon 2025) in Las Vegas, with multiple enhancements that improve support for enterprise scenarios, secure deployments, and CI/CD workflows. Users can now shape their data replication strategies with the new mirroring features.
Support for Firewalled and Secure Sources
Microsoft Fabric Mirroring now supports databases behind firewalls via On-Premises Data Gateway and Virtual Network Data Gateway, including Azure SQL Database, Snowflake, and more.
New Source Support – Azure Database for PostgreSQL Flexible Server
Database Mirroring now supports near real-time replication of PostgreSQL Flexible Server instances, adding to a growing list of sources like CosmosDB and Azure SQL Managed Instance.
CI/CD for Mirrored Databases (GA)
CI/CD support for Mirroring is now generally available, enabling teams to integrate source control with Git and manage deployments through ALM pipelines.
Workspace Monitoring Data Integration
Mirroring logs can now be tracked via Fabric Workspace Monitoring, giving teams full visibility into replication status, latency, table-level changes, and operational issues using KQL queries and Power BI dashboards.
Schema and Special Character Support
Mirroring now supports source schema preservation and Delta column mapping, allowing replication of tables with spaces or special characters in column names.
Open Mirroring Enhancements
Improvements to the Open Mirroring experience now allow users to create mirrored databases and upload Parquet and CSV files directly through the UI, making it easier to test and integrate custom sources.
Improved SQL Database Mirroring Options
New features include support for mirroring tables without primary keys and setup with reduced SQL roles, improving security and flexibility in enterprise environments.
Free Compute and Storage for Mirroring
Microsoft Fabric now offers free mirroring compute and storage (up to a defined capacity-based limit). For example, an F64 capacity comes with 64 TB of mirroring storage at no additional cost.
Regional Expansion
Mirroring is now available in additional regions, including Central US, Poland Central, and Italy North, making the feature more accessible across geographies.
These enhancements collectively make Mirroring a production-ready option for enterprise data replication, enabling real-time analytics, streamlined operations, and reduced infrastructure complexity in Microsoft Fabric.
Microsoft Fabric Community Conference 2025 introduced a powerful suite of updates aimed at improving the productivity of developers, data engineers, and platform teams working with Microsoft Fabric, alongside a variety of hands-on workshops. From pipeline automation and parameterization to migration tools and interoperability enhancements, these features underscore Microsoft’s focus on making Fabric more flexible, efficient, and developer-friendly. Attendees had the opportunity to experience these new tools firsthand through immersive sessions led by experts in the field.
Microsoft Fabric continues to evolve into a powerful platform that allows users to immerse themselves in advanced data engineering capabilities. New capabilities introduced at Microsoft Fabric Community Conference 2025 make it easier to build dynamic pipelines, automate Spark workloads, and improve resource efficiency. These features also help developers connect with their data more effectively.
A new pay-as-you-go billing model for Spark is now available in preview. This model enables dynamic resource allocation based on actual usage, offering an alternative to fixed capacity tiers — especially useful for organizations running unpredictable or ad hoc Spark workloads.
Microsoft Fabric now supports parameterized Spark job definitions within Data Factory pipelines. This enables the creation of metadata-driven pipelines that can dynamically run Spark jobs, reducing the need for duplicative workflows and improving scalability in data engineering processes.
We also found out at Microsoft Fabric Community Conference that pipelines can now be automatically triggered by file changes in OneLake (such as file creation, deletion, or renaming). This automation allows teams to build event-driven data workflows — for instance, triggering an ETL process when new data lands.
Available in preview, variable libraries let teams assign environment-specific values to variables. This makes it easier to move workloads across dev, test, and prod environments, supporting cleaner CI/CD processes.
Incremental refresh is now generally available in Dataflow Gen2. This feature enhances performance and efficiency by only processing new or changed data, which is critical for scaling dataflows in production pipelines.
Custom data engineering logic just got easier to manage. We found out at Microsoft Fabric Community Conference that Fabric User Data Functions enable teams to write once and reuse custom operations like validation and cleansing across multiple pipelines and projects.
Fabric is becoming more open and integrated with the power platform and other Microsoft technologies. With updates like support for Apache Airflow, enhanced connectors, and OneLake Shortcuts with CI/CD support, it’s easier than ever to plug Fabric into your existing workflows and tools.
Apache Airflow is now fully supported in Microsoft Fabric, enabling teams to drive productivity and efficiency by using familiar orchestration tools as part of their Fabric data pipelines.
Additionally, users can easily escape modal windows to manage workflows more effectively.
They showcased at Microsoft fabric community conference that Copy Job is now generally available, offering a faster and more intuitive way to move data across sources and into Fabric. It supports a wide range of connectors, including Oracle, Snowflake, SAP HANA, Google Cloud Storage, and more — plus ODBC sources via gateway.
Key updates include:
New event stream connectors are now in preview, alongside general availability for Azure and Fabric events integration. These enhancements improve support for real-time data ingestion and event-driven architecture.
Fabric continues to expand the utility of OneLake Shortcuts, which allow connection to external data sources (across clouds or on-premises) without copying data. CI/CD support for these shortcuts is now in preview, helping teams integrate them into automated deployment workflows.
These new capabilities, confirmed at Microsoft Fabric Community Conference, improve integration with Dynamics 365 data, making it easier for organizations to include business application data in their analytics and data engineering workflows within Fabric.
Now generally available, this feature simplifies collaboration across Microsoft 365 tenants, helping large or federated organizations securely share data and workspaces within Fabric.
Microsoft Fabric Community Conference 2025 introduced meaningful improvements to the search experience in OneLake, aimed at helping users find the data they need faster and more intuitively.
Users can now search subitems like tables, columns, and measures directly within their data items — no more drilling through layers of metadata.
This update significantly improves discoverability and speeds up workflows, particularly in complex data environments.
A new quick action feature allows users to jump directly into relevant items, making the entire search and access process more efficient.
Fabric now supports a broader range of digital development and DevOps workflows. New tools and previews empower teams to automate, extend, and manage Fabric environments using modern engineering practices.
Another interesting feature presented at Microsoft Fabric Community Conference is that developers and partners can now build custom workloads that integrate into Fabric — a foundational step for expanding platform extensibility.
A new CLI tool introduces a code-first approach to managing Fabric resources, making it easier to automate tasks and manage environments via scripts.
Updates include support for GitHub Actions using service principals and integration of variable libraries — further streamlining Fabric DevOps workflows.
To stay informed about the latest CI/CD enhancements, make sure to sign up to our newsletter for updates and notifications.
Managing streaming data solutions is now easier and more collaborative with the general availability of CI/CD tools and REST APIs for Fabric Eventstream. This update from Microsoft Fabric Community Conference brings powerful version control, deployment automation, and productivity enhancements for teams building streaming applications in Fabric.
Key capabilities include:
Fabric now supports DacFx (Data-tier Applications Framework) for managing warehouse schema changes through Git-connected deployment pipelines. With DacFx, teams can:
The Terraform provider for Microsoft Fabric is now generally available, enabling Infrastructure as Code (IaC) for Fabric environments and aligning with modern DevOps practices.
The keynotes at FabCon 2025 weren’t just incremental—they signaled a transformative leap forward for Microsoft Fabric. From enterprise-grade security with OneLake Security to democratized AI through Copilot, and from seamless migration tools to expanded CI/CD support across the stack, Fabric is rapidly maturing into a platform that’s ready for modern, scalable, and intelligent data workloads.
These updates reflect Microsoft’s clear focus on enabling organizations to build AI-ready architectures, govern data more effectively, and empower developers with the tools they need to move faster and smarter.
To dive deeper into the features and roadmap revealed at Microsoft Fabric Community Conference, we encourage you to explore the official blog posts, check out detailed walkthroughs from community experts, and join the conversation across the Fabric ecosystem.
Looking to explore what these innovations mean for your organization? Visit our Microsoft Fabric service page to see how our experts help companies unlock the full potential of Microsoft Fabric—from strategy and architecture to implementation and beyond.
.svg){kind=small}
