Moving from Power BI to data engineering

Hi, I was in a  similar situation to yours, economical / statistical background, ended up working a lot with BI tools and low-code stuff, and at some point I realized I wanted something more “technical” and flexible.

If I were you, I’d structure the transition like this:

• SQL first, honestly this is the backbone of anything data-related (especially data engineering). Getting really solid here pays off immediately.
• Then Python, focusing on pandas or polars. Don’t overdo general Python at the start, keep it very data-oriented.
• After that, move to PySpark that’s where things start to connect with tools like Databricks and real data engineering workflows.

Given you already know DAX and M, you’re not starting from zero, you already understand data modeling and transformations, which is a big advantage.

A couple of extra thoughts:

• Try to build small projects (ETL pipelines, data cleaning + storage, etc.), not just tutorials
• Get comfortable with Git and basic software practices (this is often overlooked coming from BI)
• Don’t rush into too many tools, fundamentals matter more than the stack name

Overall, your direction makes sense

Hello BIA1996 

You are in a good position already, as you have been working with DAX and M language. Your direction of travel to learn a code based approach also seems appropriate. If you've already done DP-600 your best bet is to try Notebook approach using SQL or Python language, and get some experience in working with Spark APIs. 

You can start by

- Using T-SQL with Warehouses, and get your confidence up in writing stored procedures and functions. 

- Using Python and Spark notebooks where appropriate in building ingestion, transformation and orchestration logic, Pandas first then diving into Spark. 

You background on BI is desirable - engineering can be transform the data into a desired state, but BI skills are invaluable in turning transformed and cleaned data into insights!