It's easy to want to skip straight to building impressive AI systems. It's a lot less easy to sit with a dataset and figure out what question it can actually answer.
I noticed that most of what excites me isn't the model itself — it's the process before it. Cleaning data, exploring it, finding out what's actually true about it before assuming anything. That process has a name, and it's data science, not artificial intelligence.
So I made a deliberate choice to treat data science as the foundation, not a detour on the way to something else. Statistics, structured querying, and honest exploratory analysis — the unglamorous parts — are what everything else actually stands on.
Machine learning still matters to me, and it's very much part of where this is going. But I'd rather build that on a foundation I actually understand than skip ahead and end up with skills that only work when the data behaves exactly like the tutorial version did.
This portfolio itself reflects that shift — it's less about sounding advanced, and more about being honest about where the real foundation is being built right now.