AI research engineer: mechanistic interpretability, world models, and human-aligned AGI.
How we shipped an agentic security layer for AI swarms at the UCL AI Engine hackathon, from idea to demo in one day.