Kevin Bravo

Some thoughts on the gaps we carry into harder problems, learning with LLMs, and why building someth…

Uncover system failures with LLM evaluations for AI agents, measure performance under varied conditi…