AI产品经理（评测方向）-TikTok
Core
Designing evaluation metrics and safety standards for AI agents, building automated systems to detect quality issues, and establishing ground truth datasets for AI testing.
Role type
AI Product Manager (Evaluation & Safety)
Builds
Automated quality monitoring systems, quantifiable evaluation frameworks, and ground truth datasets for AI agents.
Domain
Artificial Intelligence / Customer Service / Content Safety
Deliverable
production ML models
Required skills
Metric design and measurement methodology, Sampling and experimental evaluation, Data sensitivity and analysis, Cross-functional collaboration, Quantitative goal decomposition
Preferred skills
LLM-as-a-Judge experience, Ground Truth construction, Customer service or content safety background, Multi-region data compliance experience
Technologies
LLM-as-a-Judge, Ground Truth systems
Responsibilities
Define quantifiable standards for AI agent performance and safety, Build AI-native evaluation and ground truth systems, Design sampling methods to reflect user distribution, Establish end-to-end loops for automated quality issue detection and resolution.
