evaluated GPT-6: Astra, Sol, Luna in Codex, with a focus on agentic coding, CUA, memory/compaction, connector usage, and reasoning calibration; iterated toward a secure, reliable eval environment; helped incubate an agent-driven evaluation system for R&D. member of technical staff @ codex core agent

agi-pilled; exceptional people from many disciplines; moving fast.

explored training lower-cost subagents with multi-agent RL; improved coding agent quality through richer context and failure analysis. research scientist @ quality and cost

a great startup to learn how to run a professional tech business.

worked on early-stage LLM development for Apple Intelligence, spanning production data synthesis, agentic post-training, distillation, and eval design. research engineer @ ix-nlp

a low-key, focused vibe; room to go deep on technical details.

sequential recommendation and retrieval; AutoML research engineer @ ai research

excellent researchers and an inventive spirit.