Agent 评测算法实习生（J106616）
Core
Building an evaluation system for AI agents in industrial decision-making scenarios (scheduling, routing, bin packing, prediction) including environments, benchmarks, and platforms.
Role type
Intern, AI Agent Evaluation Engineer
Builds
Evaluation environments, standardized benchmark datasets, and evaluation platforms for industrial agents
Domain
Artificial Intelligence / Operations Research / Industrial Optimization
Deliverable
production ML models
Required skills
Python, pandas, Linux, Git, Large Language Models (LLMs), Agent frameworks, Operations Research, Machine Learning
Preferred skills
Benchmark construction, Combinatorial Optimization, OR-Tools, Gurobi, HiGHS, Kaggle/AtCoder experience, Academic publications
Technologies
Python, pandas, Linux, Git, OR-Tools, Gurobi, HiGHS, Claude Code, Codex, OpenCode
Responsibilities
Design and build evaluation environments for real-world industrial scenarios; Construct and iterate large-scale benchmark datasets; Conduct research on advanced agent capabilities (long-horizon optimization, self-evolution); Track and publish state-of-the-art benchmarks; Develop evaluation platform capabilities
Seniority
Intern
