Hey fellow data friends! Wow, March was an absolutely crazy month (in the best way). It was incredible to see so many folks in Redmond for the MVP Summit, and holy cow FabCon was a whirlwind of amazing people to meet and see from the data community. Thank you so much to everyone who came and saw my session! You folks make presenting worth it, so thank you for all the thoughtful questions and for laughing at my puns and jokes!

It was so great to see a few of my fellow P3 peers! Shout out to Garett Medlin, Jason Hurley, and Luke VanLaar for being awesome conference buddies and having awesome discussions after sessions. See below for the notes from my session and the announcements we are most excited about!

My Session – Exploring Real-Time Horizons with Microsoft Fabric

Thank you so much to everyone who attended my session! I seriously considered pulling a prank on all of you in honor of April Fool’s Day, but decided all the puns throughout the session were prank enough. Thanks for laughing and learning with me! If you missed it or would like a speed round refresher of the session, I gave a mini version of the session for Kratos BI on the Fabric Friday live stream (link below). Tune in every Friday morning on YouTube for more Fabric insights and laughs to kick off the end of the week! Would love to keep the conversation going

Link to my GitHub with slides, code, and pbix files – https://github.com/Anytsirk12/DataOnWheels/tree/main/2025%20FabCon%20Vegas

FabCon – a Red Bull Fire Hose

FabCon is the definition of drinking from a fire hose, but a firehose filled with Red Bull – it gives you wings! In an ever-evolving technical landscape, it’s vital to stay inspired. The burnout is real, especially with technologies like Fabric and Power BI that have new released daily. Conferences surrounded by 6500 fellow data friends inspire better than anything else I know. Don’t get me wrong, conferences that size can quickly be overwhelming for us introverts in the group, but there’s something special about us folks with that data gene. We speak the same language, have the same problems, and are genuinely excited to help people succeed. Theres not many places like that, so thank you to the organizers who have fostered this community and make these events a great place to invest in ourselves and others.

The Hits – Our Favorite Announcements

There were so many sessions and announcements at FabCon, but below is a list of the ones we are most excited about!

DAX

1. UDFs (DAX)
   • Coming to a data model near you, UDFs will allow semantic modelers to easily reuse code across their model (and some day across models) to limit the need to repeat code across measures! Super excited to see how this feature shapes the future of DAX coding as we get closer to following DRY principles of other languages.
2. DAX Custom Calendars
   • The possibilities are endless for this feature. We will soon have the ability to assign a column to “Year”, “Month”, “Quarter”, etc. plus some week-based time intelligence! This will enable easy use of built in time-intelligence functions with calendars like your fiscal calendar, commercial 445 calendars, or whatever calendar you prefer.

Direct Lake

1. Composite Mode for Direct Lake
   • This feature will enable us to use direct lake for our large fact tables but keep our dimensions in import! This should help drastically improve performance and follows best practices much closer.
2. Live Edit Direct Lake in Desktop
   • You can now edit your direct lake semantic models in Power BI Desktop!

Fabric Engineering

1. Variable Library
   • Enables variables to change across dev, test, prod environments
   • Can be used within pipelines, notebooks, and shortcuts in lakehouses
2. Fabric Data Factory Support for Key Vault
3. OneLake Security (yay!)
   • This demo was by far one of my favorites. So cool to see OLS and RLS respected everywhere in OneLake (a direct lake semantic model, warehouse, notebook, short-cutted lakehouse table, etc.) after only setting up the security in one spot. You’ll still need to set it up in situations like an imported semantic model where the engine isn’t actively passing credentials to the source system in OneLake.
4. User Data Functionss (Fabric)
   • Warehouse now supports SQL-based functions!
   • Azure Functions equivalent is available in Fabric. Currently only leverages python, but this is going to be a game changer. These can be called from notebooks, pipelines, and event SQL directly. These are also the mechanism for translytical task flows.
   • AI functions will allow you to use built-ins like ai_analyze_sentiment to easily perform LLM-backed columns
5. Translytical Task Flows
   • Write-back for Power BI is here! Currently, it’ll be able to pass in a set of values to a user data function (UDF in Fabric) to execute a data change.
6. Copilot for all! No more F64 requirement for Copilot!