AI评测负责人（Agent/大模型方向）
Core
Lead AI evaluation and quality assurance for Tencent PC Manager Agents and self-developed fine-tuned models, responsible for product and model effectiveness.
Role type
Senior IC AI evaluation lead (LLM/Agent)
Builds
Evaluation systems, benchmarks, and automated testing platforms for LLMs and Agents
Domain
Consumer software / Large Language Models / Multi-Agent Systems
Deliverable
production ML models
Required skills
LLM/Agent evaluation methodologies, benchmark construction, automated testing framework architecture, Python, team leadership, cross-functional collaboration
Preferred skills
Experience with SWE-bench, Terminal-Bench, OSWorld, LLM-as-Judge, data pollution detection
Technologies
Python, LLM-as-Judge, ReAct, Function Calling, Tool Use, Planning, Multi-Agent collaboration
Responsibilities
Plan and lead AI evaluation system construction covering model, Agent, and end-to-end product evaluation; Establish evaluation-driven R&D mechanisms integrating results into model iteration and CI/CD; Lead evaluation engineering platform development including dataset management and scoring engines; Drive closed-loop feedback from user traces and Bad Cases to improve models; Analyze industry benchmark trends and adapt them to business needs.
Seniority
Senior, hands-on IC with team leadership
