This article by Microsoft Fabric Blog offers a comprehensive summary of July 2025’s new features in Microsoft Fabric, authored by Patrick LeBlanc. Key updates span data science, governance, Power BI, data connectors, and AI-driven integrations with Copilot Studio.

Fabric July 2025 Feature Summary

Authored by Patrick LeBlanc, Microsoft Fabric Blog

Overview

The July 2025 update for Microsoft Fabric brings several enhancements across the platform, including feature launches, AI-driven integrations, improved real-time intelligence, enhanced data management, and upcoming community events.

Contents

  • Events and Announcements
  • Fabric Platform Enhancements
  • Data Science Features
  • Real-Time Intelligence Updates
  • Data Factory Enhancements
  • Mirroring Improvements
  • Database Updates (Cosmos DB)

Events and Announcements

Microsoft Fabric Community Conference

  • The Fabric Community Conference returns, scheduled for September 15-18, 2025, in Vienna, Austria. The event will feature full-day tutorials, a partner pre-day, and over 120 sessions from both product teams and the community.
  • Another edition will take place in Atlanta, USA, on March 16-20, 2026.

Power BI 10th Anniversary

  • Power BI’s 10th anniversary was celebrated on July 24th with community events, an on-demand birthday stream by Guy in a Cube, a Dataviz Contest, and 50% exam voucher offers for Fabric and Power BI certifications.

Fabric Platform Enhancements

Domain Tags

  • Supports organizing data into domains and sub-domains, with new capabilities allowing each domain to define and apply its own tags for business-specific contexts. This aids discoverability and governance.
  • Tags can be configured by tenants or admins and make searching and filtering content by domain more effective.

Updated Default Category in OneLake Catalog

  • OneLake catalog introduces dynamic default category assignment based on user roles:
    • Power BI users default to the ‘Insights’ category for analytics and visualization tools.
    • Fabric users default to the ‘Data’ category for raw and structured data assets.
  • User selections are preserved across sessions for a smoother experience.

Data Science Features

Fabric Data Agent Integration with Microsoft Copilot Studio

  • Fabric data agents, which are AI-powered assistants for enterprise data synthesis and policy enforcement, can now be integrated with Copilot Studio.
  • This enables organizations to build, deploy, and scale intelligent agents that work across trusted data sources, fostering agent-to-agent collaboration and richer user interactions within business chat scenarios.

Data Source Instructions

  • Fabric Data Agents now support data source-specific instructions, allowing users to guide AI behavior regarding query formulation, filtering, and dataset joins.
  • This increases precision and relevance in responses, especially in environments with complex schemas.

Streaming Results in Fabric Data Agent

  • Data Agent now supports streaming incremental query results, displaying live updates as queries are processed, thus reducing wait times and providing transparency for troubleshooting.

Improved Run Execution Flow for Data Agent

  • Enhanced visualizations and step nomenclature for each phase of a Data Agent operation, streamlining understanding and troubleshooting.

Real-Time Intelligence

Simplified Rule and Object Creation

  • Improved workflows for creating and grouping rules, reducing complexity in tracking and alerting on data events at both event and object levels.

Send Alerts to Teams Groups and Channels

  • Activator extends the ability to send automated Teams alerts not just to individuals but to group chats and channels, enhancing workflow integration and collaboration.

Pass Parameter Values to Fabric Items (Preview)

  • Activator can now pass both static and dynamic parameter values to triggered pipelines and notebooks, making events-driven automation richer and more adaptable.

Data Factory Enhancements

Incremental Copy in Copy Jobs (Generally Available)

  • Incremental copy functionality is now GA, supporting more resource-efficient data movement by syncing only changes between runs, reducing processing and costs.

Upsert Data to More Destinations

  • Copy jobs can now upsert (merge) directly into Fabric Lakehouse, Salesforce, Dataverse, Dynamics 365/CRM, and Azure Cosmos DB (NoSQL).

Expanded Connectors

  • New source/destination connectors include SFTP, FTP, IBM Db2, Oracle Cloud Storage, Azure AI Search, and many more.

Copy from On-Premises to Data Warehouses

  • Copy jobs now natively support copying data from on-premises data stores directly into Snowflake and Fabric Data Warehouse using staging in OneLake.

Manual Control of Auto-Refresh in Pipelines

  • Users can now disable or enable auto-refresh of output and monitoring views in pipeline activities, allowing easier exploration of lengthy activity lists.

Mirroring Improvements

Azure SQL Database Mirroring (Firewall Support, GA)

  • Now generally available with support for network security via VNet Data Gateway and On-Premises Data Gateway. Enables secure mirroring of Azure SQL Databases behind firewalls.

Improved Resume Process for Mirrored Databases

  • When Fabric capacity is resumed, mirrored databases accurately show their ‘Paused’ status and allow manual resume, maintaining data integrity across interruptions.

UI Option for Retention Period

  • The UI now supports direct configuration of mirrored data retention periods, offering more flexibility over data lifecycle management.

Database Updates

Cosmos DB (NoSQL) in Fabric (Preview)

  • Azure Cosmos DB (NoSQL) is now available in preview for Microsoft Fabric, enabling direct integration and real-time analytics on globally distributed NoSQL data within Lakehouses, Notebooks, and Power BI.
  • This reduces ETL workload and supports development of AI-driven applications on a unified platform.
  • OneLake as a Source for COPY INTO and OPENROWSET (Preview): Simplifies data ingestion for Data Warehouses.
  • JSON Lines Support in OPENROWSET: Adds parsing capabilities for semi-structured data in Data Warehouses and Lakehouse SQL endpoints.

Further Reading and Documentation

For more detailed instructions on each new feature, view the referenced documentation or follow the embedded links for tutorials and community resources.

This post appeared first on “Microsoft Fabric Blog”. Read the entire article here