← Back to Careers

Algorithms & Software · Full-Time

LLM Algorithm Engineer

BeijingXi'an

What You Will Do

  • Build LLM agents: tool use, multi-step reasoning, task decomposition, workflow orchestration; optimize RAG, memory, planning, reflection, multi-agent collaboration; drive engineering deployment.
  • Enhance foundation models: reasoning, knowledge injection, context management, alignment; explore RL/CV fusion; track research and feasibility.
  • Produce docs/patents/papers; join tech sharing; support team and cross-team collaboration.

What We Are Looking For

  • Bachelor's+ in CS, Math, Statistics, or related; 3+ years in algorithm or deep learning.
  • Solid ML/DL foundation; understanding of Transformer, LLM inference, and alignment.
  • Proficient in Python/C++; strong engineering and system design skills.
  • Familiar with LLM inference/deployment: LoRA/QLoRA, quantization, mixed precision, memory and communication optimization.
  • Familiar with Agent frameworks: LangChain, LangGraph, Function Calling, ReAct, etc.; hands-on tool use, multi-step orchestration, and workflow deployment.