Product Manager Leader (GenAI Safety Evaluation) - TikTok Safety Product (San Jose)
TikTok
Posted 2 days ago
About this role
About the Team
The Safety Product (or Platform Responsibility) team is at the forefront of building and optimizing content safety systems. With a focus on optimising and advancing content safety, we leverage advanced large language models to enhance review efficiency, risk control, and user trust. Working closely with business and technical stakeholders, we deliver scalable solutions that keep pace with rapid global growth.
Responsibilities:
- Own and evolve the GenAI / LLM safety evaluation framework across multimodal use cases, including text-to-text, text-to-image, and text-to-video, with a focus on measuring model performance across core safety and risk dimensions.
- Design and scale online safety evaluation systems, introducing new evaluation methodologies and driving automation and large-scale deployment to enable continuous model and product iteration.
- Build and maintain safety benchmarks by creating high-quality evaluation datasets and adversarial test cases, ensuring benchmark distributions closely reflect real-world user behavior and risk scenarios.
- Establish and continuously improve root cause analysis (RCA) processes, including standardized workflows for issue detection, attribution, and postmortem reviews, and distill reusable risk patterns and actionable optimization insights.
- Partner cross-functionally with Model, Algorithm, Product, Data, and Safety teams to translate key safety findings into optimization roadmaps and ensure closed-loop validation of improvements.
The Safety Product (or Platform Responsibility) team is at the forefront of building and optimizing content safety systems. With a focus on optimising and advancing content safety, we leverage advanced large language models to enhance review efficiency, risk control, and user trust. Working closely with business and technical stakeholders, we deliver scalable solutions that keep pace with rapid global growth.
Responsibilities:
- Own and evolve the GenAI / LLM safety evaluation framework across multimodal use cases, including text-to-text, text-to-image, and text-to-video, with a focus on measuring model performance across core safety and risk dimensions.
- Design and scale online safety evaluation systems, introducing new evaluation methodologies and driving automation and large-scale deployment to enable continuous model and product iteration.
- Build and maintain safety benchmarks by creating high-quality evaluation datasets and adversarial test cases, ensuring benchmark distributions closely reflect real-world user behavior and risk scenarios.
- Establish and continuously improve root cause analysis (RCA) processes, including standardized workflows for issue detection, attribution, and postmortem reviews, and distill reusable risk patterns and actionable optimization insights.
- Partner cross-functionally with Model, Algorithm, Product, Data, and Safety teams to translate key safety findings into optimization roadmaps and ensure closed-loop validation of improvements.