The Python Data Science Stack That Finally Made Me Enjoy (ML).

I rebuilt my entire ML workflow from scratch — and everything just clicked.

Press enter or click to view image in full size

The Moment I Realized My ML Workflow Was Holding Me Back

It started with a single failed notebook.

You know that feeling when your model finally runs and the results make zero sense?
That was me staring at my screen, 40 minutes deep into debugging a pipeline that felt more like duct tape than data science.

I had learned all the “right” libraries, pandas, scikit-learn, matplotlib, and NumPy, but I didn’t enjoy the process. My workflow was scattered, my environment fragile, and every small experiment meant hours of setup. It felt like the opposite of what machine learning was supposed to be: automated intelligence.

So I wiped everything. Rebuilt my data science stack from scratch.
And this time, I didn’t just chase performance, I chased clarity, speed, and automation.

What came out of that rebuild changed how I work forever.

Step 1: Start with a Stack That Automates the Boring Stuff