汽车行业Agent评测产品经理 - 火山引擎
Core
Build and execute daily evaluation systems for automotive Agents, focusing on E2E task coverage from user expression to system execution and result validation.
Role type
Product Manager (Automotive Agent Evaluation)
Builds
Stable evaluation datasets, regression test cases, and automated evaluation pipelines for automotive AI agents.
Domain
Automotive industry + Large Language Models (LLM) & AI Agents
Deliverable
production ML models
Required skills
Product management, E2E task design, data analysis (Excel/SQL), Python/AI coding, LLM/Agent mechanism understanding, issue tracking and closure.
Preferred skills
Experience in smart cars, smart cockpit, vehicle control, navigation, multimedia, automated evaluation, user blind testing.
Responsibilities
Design and maintain E2E evaluation cases covering core scenarios like vehicle control and navigation; extract high-value scenarios from user feedback and online data; convert historical negative cases into regression tests; execute version evaluations to validate full task completion; analyze evaluation results to identify issues in intent understanding, reasoning, tool calling, and execution; participate in building automated evaluation capabilities.
Seniority
Mid-level, hands-on IC
