Senior Product Manager, AI Safety Evaluation & Governance - TikTok Safety Product

Written by the T&S Network from TikTok's posting.

Defines and runs the evaluation system that decides whether AI models are good enough to make content safety decisions across the platform. The work covers standards, model design and statistically grounded quality measurement for safety models, sitting within the Feed Safety-Model and Data Intelligence team in TikTok Platform Responsibility.

What you'd do

• Define evaluation methodology, launch criteria, regression thresholds and monitoring for AI safety models
• Build labeling taxonomies and positive-example datasets across languages, formats and adversarial patterns
• Create severity-tiered benchmarks to set quality gates and calibrate decision thresholds
• Compute precision, recall, F1 and confidence intervals, and identify model failure patterns
• Write and tune prompts for LLM-as-judge and LLM-assisted labeling pipelines
• Coordinate labeling teams and translate findings into roadmaps with algorithm, policy and data science teams
• Build reusable evaluation tooling and mentor junior team members

What it asks for

• Degree at bachelor's level or higher in a quantitative field such as statistics, computing or data science
• 5+ years in AI/ML evaluation, model QA, trust and safety product, or data-driven product management
• Strong SQL for statistical analysis and data pipeline work
• Proven professional experience with LLMs, including prompt engineering and LLM-based evaluation
• Deep understanding of content safety or trust and safety policy, severity classification and enforcement trade-offs
• Fluent written and spoken Chinese for managing China market stakeholders
• Has built evaluation frameworks for large content moderation or safety AI systems (preferred)

Our best summary of the posting, and it can be wrong. The employer's posting is what counts: read it in full on TikToknorth_east before you apply. If we got something wrong, report this listing.