策略算法工程师
Core
Design and implement core capabilities for intelligent agent Harness Engineering, including agent loops, multi-agent collaboration, memory, skills, task planning, tool use, and state management to enhance stability and task completion quality.
Role type
Senior IC strategy algorithm engineer (intelligent agents)
Builds
High-concurrency agent service frameworks, scheduling systems, and workflow engines for enterprise scenarios
Domain
Artificial Intelligence / Large Language Models / Enterprise Software
Deliverable
production ML models
Required skills
Go, Python, high-concurrency system design, distributed services, task scheduling, workflow engines, LLM function calling, tool use, ReAct, Plan-and-Execute, agent evaluation metrics
Preferred skills
RAG, instruction tuning, model evaluation, multi-agent collaboration, bad case analysis, enterprise product delivery
Technologies
Go, Python, LangGraph, AutoGen, Coze, Claude Code, Codex, OpenCode, MCP
Responsibilities
Design and implement core agent modules (loop, memory, planning, tool use); Optimize prompts for accuracy and robustness in complex enterprise scenarios; Build high-concurrency agent service frameworks and scheduling systems; Establish evaluation and observability systems for agent performance; Analyze bad cases to drive iterative optimization; Collaborate with product and business teams to deliver high-quality agent capabilities.
Seniority
Senior, hands-on IC
