Justin Huang

Back

Contact

static post

Happy to talk about coding agents, Agent Infra, LLM post-training, scientific tooling, AI-native workflows, or this site itself. I may reply slowly, but I do read carefully.

How to reach me

What to talk about

  • Agents / Coding Agents: tool protocols, sandboxes, verifiers, long-horizon tasks, and SFT/RL data quality.
  • AI Infra: runtimes, gateways, memory, observability, prompt evaluation, billing, and failure recovery.
  • Research and writing: Deploy-Master, XRAG, SAE, Terminal-Bench, and personal knowledge systems.