Nlp, Python, Ml, Rust, prompt caching, chain-of-thought reasoning, RLAIF, preference data construction, long-context training, QAT, DPO, MoE, Go, RLVR, Distillation, KV-cache, online RL, applied DL, TensorRT-LLM, Agent engineering, continual pre-training, LLM-as-judge, vLLM, reward modeling, multi-node SFT

Qt, Code Review, Mcitp, Nlp, Roadmap, Api, Python, Distillation, Failure Analysis, data consistency, Agent Management System, Technical Design, Liaise with design team