Back to Home

My Thoughts on SQL Conference 2025

Published on May 11, 2025 by Fabian Stadler

It’s already been some weeks but at the end of February, I could attend this year’s SQL Conference in Hanau. For the past months, I’ve been spending more time to skill up on the newest stack of data engineering technologies, one of them being Microsoft Fabric. And as this year’s conference was also having lots of sessions on this, I took part from our Microsoft center of excellence at HMS Analytical Software.

Unlike the name suggests, the SQL Conference is mostly not about SQL. It’s a conference led by Datamonster e.V., an association of data enthusiasts across Germany. And Microsoft enlisted quite a lot known speakers, among others. Therefore, and as the community revolving around this conference also comes with a strong background on SQL Server, the topics are pretty much on Microsoft technologies.

What I found pretty interesting was a session by Stefan Kirner, Director Business Intelligence at scieneers GmbH, who talked about migrating Azure Data Factory (Azure Synapse Analytics) to Microsoft Fabric. It astonished me that there is no proper solution yet to migrate simple things like data pipelines or linked services (connections to third party systems) even though there are similar counterparts in Fabric. Worse, even though you have a repository with code-based configuration in ADF, you cannot migrate because Fabric is using GUIDs and references in pipelines.

Many times, especially when there is no counterpart in Fabric, like for Mapping Data Flows, you’re left to convert elements into Spark notebooks that you can then adjust and load into Fabric notebooks. I hope Microsoft will soon or later bring some tooling to make this straightforward. Yet, the challenging and complex migration process could also be an opportunity that I will try to keep my eyes on.

Pieter Vanhove, Program Manager in the Security & Governance Azure Database Platform team at Microsoft, showed how to use secure enclaves with Azure SQL Database to process non-deterministically encrypted data. Without losing the ability to perform more complex SQL operations, leveraging Intel’s SGX or the VBS platform. These provide physically isolated memory and processing capacities that can temporarily decrypt and process encrypted data. However, the attestation process to rule out manipulation looks complex to set up. And I also did not yet work on a project that required such strict data security measures. So, I look forward if there will be an opportunity in the future.

Last but not least, there were some other, sometimes funny, presentations that I enjoyed, such as a Power BI dashboard with a similarity search for Lego-sets based on embeddings. Or a deep-dive into Fabric’s real-time-intelligence that showcased the potential of Fabric with Power BI real-time dashboards. Especially the latter inspired me a lot to find a good use case for a test project.

While there were lots of good sessions, most of them took place on the first day. The second day could have been more interesting. Since I also haven’t worked with SQL Server and related technology much, some sessions focussed too much on DBAs while others I looked forward to were quite basic. All in all, I think it was worth visiting, especially for the food, networking and, of course, the inspiration.

Links

If you have any questions or feedback, feel free to write me a mail or reach out to me on any of my social media channels.