2027AIDU-Agent应用全栈工程师(J99974)
Core
Build next-generation Autonomous Agent systems that evolve AI from chat tools into task-executing agents, focusing on complex task planning, execution, and reflection loops.
Role type
Senior Full-Stack Engineer (AI Agent Systems)
Builds
Autonomous Agent systems with tool/API calling, memory mechanisms, multi-agent collaboration, and RAG-fused architectures for real-world applications.
Domain
Artificial Intelligence / Large Language Models / Autonomous Agents
Deliverable
production ML models
Required skills
Large Language Model (LLM) development, Agent framework implementation (LangChain, LlamaIndex, AutoGen, CrewAI), Tool/API integration, Memory management, Multi-agent system design, RAG architecture, Evaluation体系建设 (success rate, stability, cost, latency), Rapid prototyping and experimentation.
Preferred skills
Experience with Self-Improvement, World Models, Multimodal Agents, Tool Learning.
Responsibilities
Design complex task planning-execution-reflection loops; Build core Agent capabilities including tool calling and state management; Explore and implement Multi-Agent Systems for business automation; Participate in RAG+Agent architecture design; Establish evaluation systems to optimize agent performance and cost; Track and engineer frontier directions like Self-Improvement and World Models.