Seven reliability primitives for any LLM API: Chain-of-Verification, self-consistency, best-of-N, reflexion, debate, drift detection, hedged execution.
Local-first personal AI assistant — CLI, web, and desktop frontends over a shared agent core.
A pre-spend GATE + learning advisor for LLM API cost: caps every call, prices from a verified table, and learns the cheapest config that keeps quality.
To truly close the digital equity gap, educators must become architects of student possibility.
An evaluation framework for VLA (vision-language-action) models across real robots and simulators — the Inspect AI for robotics.
