Amine Azariz

Most of my career has been in fintech. Payments, open banking, regulated systems that handle real money for real businesses. The kind of software where a wrong assumption costs more than a bad sprint.

That background is why PyStack looks the way it does. When you've spent years dealing with systems that have to be right, you get skeptical of anything that only works cleanly in demos. The picks on this site reflect that.

What I'm working on now

Finance is one of the most underused areas for applied AI right now. Most teams got as far as a support chatbot and stopped. The genuinely interesting work is in the data underneath, specifically in transaction data and how money actually moves.

Right now, that means building ML models that classify business transactions from open-banking data. There's been a lot of work done on the consumer side, categorizing personal spending, making sense of someone's checking account. The business side is a different problem. The data is messier, the categories are harder to define, and there's far less established tooling. That's exactly what makes it worth doing.

PyStack grew out of this work. When you're evaluating Python libraries against a real problem, not a benchmark, you develop strong opinions fast. The tools that made the list here are tools that have held up under that kind of pressure.

The curation process is described in more detail on the Why page, if you want to understand the methodology. The short version: production use, active maintenance, and a clear job to do. Anything that only shines in a README doesn't make it.

If a pick looks wrong to you, or something's missing: LinkedIn.