
Content Risk Evaluation Team Lead (Mandarin Speaking) - Safety Model Operations - Kuala Lumpur
Job description
About the team
The Safety Model Operations [SMO] team is responsible for building, optimizing, and maintaining machine learning models and operational processes that support TikTok's Trust & Safety systems. We ensure that automated safety models perform effectively in identifying harmful content, mitigating risks, and maintaining a safe user environment across regions.
The SMO Delivery Team plays a critical role in the organization. Its primary responsibility is to carry out the full spectrum of quality assurance activities for each project. This includes:
- Conducting detailed reviews and complex RCA's to ensure labeling accuracy and consistency
- Monitoring quality performance and compliance against project-specific KPIs
- Identifying trends, risks, and potential gaps in processes or guidelines
- Providing structured feedback and improvement recommendations to the Central Project Team
- Supporting continual optimization of workflows, tools, and evaluation methodologies
- Improve Model performance of AI models
What will I do?
As the Content Risk Evaluation Team Lead, you will build and lead a team of Content Risk Evaluation Specialists who serve as the frontline guardians of the platform's safety ecosystem. You will guide your team in translating complex regulations and community standards into actionable guidelines, proactively surfacing hidden and borderline content risks across both user and producer-generated content. When critical risk events emerge, you will orchestrate rapid, cross-functional responses to contain and resolve threats while continuously improving review processes. Your team will produce data-driven risk assessment reports that directly shape platform rules and safety strategies. Additionally, you will work hand-in-hand with algorithm and data science teams to close the feedback loop between human judgment and AI models—ensuring consistency, reducing model edge cases, and driving continuous improvement in both human review quality and automated content safety systems.