I got here the long way round. I was pointed at the rigorous, prestigious academic path early, and at the start it did not interest me one bit.
That changed about a year ago, once I noticed the work I actually finished was always work on a problem that was bothering me personally. That is still the whole argument: if I can solve a problem for myself, why not optimise the process and solve it for everyone else too. Most of what is below started exactly that way.
Learning all of this in the middle of the AI era taught me the other half. In a world full of frontier models you can certainly build cool things by outsourcing the intelligence, and it is tempting. Forcing myself to take the extra mile and learn the thing while I build it is what makes the process exponentially better.
Right now I am working through ML fundamentals, RAG evaluation, computer vision, and what it takes to run AI in production. I am looking for an AI engineering role.
Cards marked live demo open the demo; the rest open their repository.
Weekly minutes from my Hevy log over the last 16 weeks. The flat stubs are weeks I did not train.
I am open to AI engineering roles and to collaborating on applied ML and trustworthy AI. If you have a hard, practical problem, send it over.
Email me or connect with me on LinkedIn.
