Microsoft Fabric’s Next Leap: Insights from FabCon Europe 2025

FabCon25 Europe or, to give it its full name – The European Microsoft Fabric Community Conference is the European version of Microsoft’s annual data and analytics platform conference. This year it was held in Vienna and attracted over 4,000 attendees. It is now the largest Microsoft run conference outside of the US.

This year I was lucky enough to be invited to the conference to present a session titled ‘(Almost) Everything You Wanted To Know About Purview Data Governance’ taking lessons learned from my work at Simpson Associates with a UK regulatory body and compiling a beginning to end run through of features focusing on avoiding the traps you can fall in to that might either require re-work or limit the success of your data governance solution.

The session was well received with 100 people in the room. A quick poll of attendees showed the majority were either preparing to roll out Purview in the imminent future or evaluating it as their governance tool of choice. Conversations afterwards were great; data governance can often be overlooked so sharing experiences with others who get how fundamental it is to getting full value from your data was a joy.

Announcements from the European FabCon 2025

The most prominent feature of any Microsoft conference is the announcements. Microsoft loves saving up new features or improvements to old ones and sharing them on the big stage at their keynote sessions. This year, Microsoft had so much to cover that they needed not one but two keynote sessions to cover it all. One looking at the core analytical offerings from Fabric and another focusing on converting Fabric from being your core data analytics platform to being your core data platform full stop.

You can read the full list of features released here: Fabric September 2025 Feature Summary | Microsoft Fabric Blog | Microsoft Fabric and videos of the keynote sessions should be available to watch online soon. Here were some of the highlights from my perspective:

Git Integration and Deployment Pipelines

It’s fair to say that at release Fabric’s support for deployment between development, testing and production was somewhat lacking, with many features unsupported and complex solutions needing heavy manual or scripted interventions. Microsoft have worked hard to fill in the gaps and as of now deployments are out of preview with 100% coverage on all features. Tight integration with per workspace variable libraries allow for dynamic deployments and customisations. For those still deploying via code, command-line features continue to evolve, now with the Fabric CLI open-sourced, enabling community-driven development.

Graph Capabilities

One of the most exciting announcements for me was the introduction of Graph capabilities in Fabric. Graph databases excel at mapping relationships and properties between entities, think of datasets like IMDB, where films are linked to people in various roles. Working in the public sector, where data often follows similar patterns, I’m excited to explore and experiment with this powerful new feature.

Graph capabilities could enable analysis of relationships between families, social care workers and other key contacts as well as the flow of contacts, assessments, referrals and other events in a typical hospital stay. Copilot support is built in to the interface with an impressive ability to support querying the data think something like “Please give me a list of all Oscar winning actors who have appeared in more than five films but never alongside Kevin Bacon”.

Everyone loves a good map visualisation and Fabcon delivered with an extended demo of Fabric’s new Map feature. Bringing geospatial visualisation to all Fabric users, the map tool provides an intuitive interface that supports everything from simple heat maps to complex visualisations. The feature blends layers of geographical data such as maps and custom route information with real time overlays of activity data. Maps are expected to leave their private preview in the next few weeks.

The Bigger Picture for Microsoft Fabric

Underlying all of this is the big picture for Fabric: to become not just an organisation’s central source of analytics but to become the core of all data within an organisations. That vision is now falling in to place.

OneLake becoming a central data store not just for Fabric but supporting integration with partner products such as Snowflake. The tools to manage OneLake particularly in terms of managing access have been improved allowing management at scale.
Support for workloads delivered by partner companies. Fabric has ceased being a purely Microsoft created offering with the ability to slot in features such as master data management tools with tight integration to existing features.
Purview Data Governance is steadily improving and becoming more and more critical to Fabric’s overall vision. As the scale of Fabric grows so does the need to govern data in it and to ensure that it can be discovered and understood not just by human staff but also by the ever growing AI based offerings. Customisable metadata for all assets is now in preview and customisable workflows with integration to Power Apps is on the roadmap to be delivered over the next two quarters.
“Reverse ETL”. Typically data flows in one direction from source systems in to the data warehouse. Fabric now has a growing set of methods to allow data to flow back to source systems. Imagine you’ve created a sophisticated machine learning model in Fabric that takes inputs from multiple sources and predicts product sales over the coming days or weeks. Now imagine working to avoid shortages or over ordering, instead of having to cross reference these results between a report and your stock control system, the needed are pushed directly back to source.
AI Foundry to pull it all together generating custom chat bots that feed from data, metadata and security models that allow users to “chat with their data” (but only the data they’re allowed to see).

Conclusion

There was much much, more (Data Flows got four times faster and eight times cheaper, DAX UDFs are being described as the most significant change to the language in half a decade, MCP servers that allow Copilot to build whole reports from a specification document) and I’m sure, plenty that I missed in the post keynote break out sessions. It’s a lot to take in and hard to keep up but that’s why we at Simpson Associates are there to support you in your journey.

FabCon Europe will be back next year from 28^th September to 1^st October in Barcelona. I will be speaking at the PASS Summit in Seattle which will run between 19th and 21st November.

Blog Author:
Barney Lawrence, Senior Consultant at Simpson Associates.