Power BI and Fabric - April 25 Preview Features

As we go into May, last months update list has landed and it's time to go through what's new and what's worth knowing about. Given the volume of updates we're getting now, I'm going to keep splitting out my thoughts into:


With all things in public preview, don't forget the general rule of thumb is to not use these in production. Sometimes you have to, but do so at your own risk. Otherwise it's worth keeping an eye on to help shape roadmaps.

First up, Power BI.

Power BI 

Modelling

Live edit of Direct Lake semantic models in Power BI desktop

Currently BI devs need to build and maintain direct lake semantic models within the service. This update allows them to go back to using the more familiar approach of using PBI desktop. 

For those already using this feature, checkout the PBI blog for what the changes are. Personally, this is a feature I might consider turning on soon, at the moment the limitations in the learn documentation  aren't show stoppers for me - but you have to have the bi-directional reflection of changes working before I'd use it in anger.

Semantic model changes with TMDL

For those that are starting to learn TMDL, we've seen improvements to the UI in order to make it easier to understand the differences between different model versions. 

Data connectivity

This month we've seen a number of improvements to connectors for Vertica, Oracle and Snowflake.


Fabric

Data science

Notebook AI tools

This month we've had AI tools tab added to the notebook view. This contains all the tools needed to accelerate productivity with AI capabilities (e.g. Copilot, Data Wrangler, AI functions, etc).

Data wrangler Low-code AI tools

The following AI tools have been added to data wrangler this month These are:
  • Automated suggestions. Uses AI to highlight the most relevant operations based on the supplied data set.
  • Convert natural language to code with Copilot. Reducing the entry point for using Copilot
  • Use AI to translate custom code from pandas to Pyspark. Helping to improve performance

Datawarehouse

Migration assistant

This is aimed at those using dedicated SQL pools in Synapse. It has been designed to migrate metadata and data, but doesn't migrate any application code.

Personally, I'm on the fence about this. With typical migration patterns not being a one and done approach, I think this is too simple for a lot of use cases. However, if you have very simple requirements this could be good for you.

Be warned that this is brand new this month and is more likely to have critical bugs in it. If you decided to use it, remember to test, test, test.

Databases

With Fabric databases still in preview, we've had a number of updates this month. Including:
  • Availability in new regions
  • The announcement of backup billing
  • Performance dashboard
  • Terraform, rest API, & CLI support
  • Integrations via data pipelines
  • Graph database support
I know some people are early adopters and have started using Fabric databases. If this is you make sure you check out the blog articles

Data Factory

The only update this month is changes to Snowflake mirroring to add data gateway support. Meaning we can now access Snowflake instances in other CSPs (especially AWS), and behind VLAN's.

Comments

Popular posts from this blog

Ignite 2024 - What's been announced for Microsoft Fabric

Workspace topologies in Microsoft Fabric

Power BI - Fabcon keynote, preview features, and March 2025 announcements