微信-大模型(Agent)应用开发后台工程师
Core
Design and develop backend architecture and core systems for WeChat intelligent agent applications, enabling complex task planning, tool invocation, and multi-turn decision-making.
Role type
Senior IC backend engineer (LLM agent systems)
Builds
High-performance, high-availability agent engines, multi-turn dialogue systems, and multimodal interaction pipelines for hundreds of millions of users.
Domain
Consumer social media / Large Language Models / Agent Engineering
Deliverable
production ML models
Required skills
C++, system design, RAG, Function Calling, task planning, tool invocation, LangChain, LangGraph, Dify, Coze, database, message queues, caching, containerization
Preferred skills
model fine-tuning, inference optimization, cost management, multi-agent collaboration
Responsibilities
Design backend architecture for agent applications; implement high-concurrency, low-latency agent systems; explore and implement advanced engineering practices for agent components and model optimization.
