This is my notebook for learning AI engineering in public. I build things, run experiments, and write down what I learn.
I’m interested in what survives contact with a real task: how tools differ, where systems fail, and which trade-offs actually change a decision.
What belongs here
Focused experiments, reproducible comparisons, and technical explanations with enough detail to use. An article should leave you with a clearer decision or something concrete to try.
What doesn’t
Vendor talking points, inflated claims, or benchmarks without a method. Results will be tied to the versions, datasets, and conditions tested, not presented as universal winners.
Follow the work
Use the RSS feed to receive new posts in your reader. No mailing list or tracking cookies.