Why I Use Power BI and Python in the Same Project

The question comes up a lot: do you use Power BI or Python?

Both. Always both. And I'm tired of pretending this is a controversial take.

The Split

Python handles the messy parts:

• Scraping and ingesting data from sources that have no API
• Cleaning and normalizing before anything touches a model
• Scheduled jobs that run on the VPS at 3am without complaining
• Anything involving text, ML, or custom logic

Power BI handles the visible parts:

• Stakeholder communication
• KPI monitoring with drill-through
• Executive reports that someone will print and bring to a meeting

The mistake is trying to make one tool do both jobs.

The Real Skill Is the Bridge

DAX is underrated. Not because it's pleasant to write — it isn't — but because it forces you to think about how data should be shaped before you can express what you want.

I've learned more about data modeling from writing bad DAX than from any course.

Lessons from the Field

At CPFL and AB InBev, the most impactful dashboards were never the most complex ones.

The ones that actually changed decisions were the ones where someone looked at the screen and immediately knew what to do next.

That's the job.