企业微信-大模型算法工程师-Agent记忆系统（北京/成都）
Core
Design and implement core AI systems for WeChat Work, specifically LLM/Agent memory systems, context engineering, and knowledge graph construction to support personalized cross-session interactions.
Role type
Senior IC machine-learning engineer (Agent memory & self-evolution)
Builds
Production memory systems, context management modules, and self-evolving Agent capabilities for enterprise chat scenarios.
Domain
Enterprise communication platforms + Large Language Models + Knowledge Graphs
Deliverable
production ML models
Required skills
Transformer architectures, RAG, Knowledge Graphs, Context engineering, Python, Deep learning frameworks, System abstraction, Engineering optimization
Preferred skills
Agent self-evolution, Skill scheduling, Conflict resolution, Paper publication in top-tier conferences (ICLR, NeurIPS, ICML, ACL, EMNLP)
Responsibilities
Design and implement LLM/Agent memory systems and context engineering; Compress and extract knowledge from structured/unstructured data; Build hierarchical memory architectures with lifecycle management; Design and implement Agent self-evolution systems; Integrate memory retrieval and graph reasoning into online Agent systems.
Seniority
Senior, hands-on IC