The TikTok AI Data Service and Operations (ADSO) team advances AI across TikTok by turning human expertise and real-world insight into data for model training, evaluation, and improvement. We work with product and algorithm teams to develop data standards, build evaluation datasets, and identify opportunities to strengthen model performance.
Human judgment is nuanced and context-dependent, while models and product requirements continue to evolve. We translate that judgment into consistent data across languages, domains, and product scenarios, combining domain expertise, AI-powered automation, and global operations to improve evaluation methods and data production systems.
You will collaborate with algorithm, product, engineering, and operations teams throughout the model development lifecycle. This provides opportunities to deepen your understanding of AI systems, shape how models are trained and evaluated, and see how stronger data and evaluation practices inform product improvements.
About the Team Generative AI and large language models are reshaping the core capabilities of global content platforms — and high-quality training and evaluation data is one of the key factors that determines a model's ceiling.
We are the AI Data Service & Operations team behind TikTok's international products, responsible for producing multilingual, multimodal data assets across both safety and non-safety domains. The standards we set and the data we deliver power two critical fronts: TikTok's global content ecosystem governance strategy on one side, and the training and evaluation pipelines for AI/large language models on the other. In short, we produce the fuel that helps models truly understand content, communities, and users around the world.
We are looking for talented individuals to join our team. As a graduate, you will get opportunities to pursue bold ideas, tackle complex challenges, and unlock limitless growth. Successful candidates must be able to commit to an onboarding date by the end of the year. Please state your availability and graduation date clearly in your resume. Candidates can apply to a maximum of two positions and will be considered for jobs in the order you apply. The application limit is applicable to our Company and its affiliates' jobs globally. Applications will be reviewed on a rolling basis - we encourage you to apply early.
Responsibilities
- Help define and continuously iterate data-quality standards, judgment rules, and acceptance criteria for foundation-model training and evaluation.
- Review and judge data outputs against quality standards, identify issues, close correction loops, and ensure accuracy and consistency.
- Participate deeply in the adjudication and discussion of complex, ambiguous, and borderline cases; convert conclusions into reusable judgment rules and knowledge assets.
- Track quality metrics, identify root causes, and drive improvements in production processes and execution.
- Collaborate with production, product, algorithm, and strategy teams to align quality standards and connect the “standards–production–acceptance–feedback” chain.
Minimum Qualification(s)
- Bachelor's degree
- Excellent English listening, speaking, reading, and writing skills, including accurate comprehension of English video and text. IELTS 7.5, TOEFL 105, or equivalent.
- Familiarity with Excel and other common data-processing tools
- Clear logic and strong structured decomposition, abstraction, synthesis, judgment, and communication skills. Detail-oriented, with an independent view of what constitutes high-quality data.
Preferred Qualification(s)
- Majors in law, philosophy, economics, journalism, languages, social sciences, or related fields
- Internship or project experience in data quality, content moderation, compliance, content operations, AI data, or annotation
- Deep experience with mainstream foundation-model products such as ChatGPT, Claude, Gemini, or Doubao, and a basic understanding of AI capability boundaries