Aleksei Prikhodko
Senior AI engineer at Artificial Agency — Canada
I build AI agents that have to work in production, not in a demo: system design, evaluation, training, and the optimisation that makes them worth their cost. Before Artificial Agency, real-time agents shipped inside games at Inworld. Before that, inverse problems in applied mathematics — which is where the habit of taking things apart comes from.
Writingall →
01 Inference optimization decoded: KV cache, Flash Attention, and why your LLM is slower than it should be
The optimizations every team names but few can order by payoff. Walked through in the sequence they actually matter, with the cost model behind each one.
Builtall →
Python CanonWhy the language is the way it is: 21 lessons that trace everyday Python back to a handful of design decisions.
AP Lab AcademyInteractive courses for engineers who want to understand, not just watch: no video lectures, visualisations you can poke at.