TikTok

AIGT Agent Operations (Model Data Operations), TikTok Platform Safety

TikTok

Singapore · Full Time

Be the first to apply

Experience
3+ yrs
Salary
Openings
1
Posted
1 week ago
Work mode
In office
Education
Bachelor's degree
Eligibility
Bachelor’s degree or above with at least 3 years of relevant experience in content risk control, strategy operations, search suggestion strategy, or AI product operations; suitable for candidates with practical prompt engineering and data analysis experience who can operate effectively in fast-pace…
Resume
Required to apply

Where you'll work

Job description

About the Team

The Platform Responsibility group is focused on advancing content safety through applied AI. The team is building the next generation of infrastructure for risk detection and governance by combining strong engineering foundations, large language models, and high-quality data pipelines. Its work supports more precise enforcement, greater efficiency, and the trust that underpins global product growth.

About the Role

This position sits within the TikTok Community Guidelines framework and centers on using LLMs and prompt engineering to power an agent designed for highly accurate content safety evaluation. The role will own strategy refinement and prompt optimization work that improves the agent’s ability to detect risky content with stronger recall and precision.

Core Responsibilities

  • Create, test, and refine prompts for defined risk scenarios, while keeping prompt versions organized so the best instruction set can be applied to each use case.
  • Build and maintain a curated library of few-shot examples, making sure the sample set reflects the newest underground-industry patterns and difficult edge cases.
  • Investigate false negatives and false positives in depth to identify whether issues are caused by unclear prompt guidance, stale examples, or policy ambiguity.
  • Respond quickly to sudden public-opinion shifts or new attack patterns by adapting interception strategy, then turn those temporary measures into durable improvements after review.
  • Update and manage the external knowledge base used by the RAG system for political, current-affairs, and person-related risk areas so the agent can recognize the latest terms and language.
  • Track operational metrics every day, prepare daily and weekly reports, and alert the team promptly when abnormal trends appear so root causes can be investigated.

Qualifications

  • Bachelor’s degree or higher, plus at least 3 years of experience in content risk control, strategy operations, search suggestion strategy, or AI product operations.
  • Strong understanding of content safety workflows and related business requirements.
  • Hands-on knowledge of LLM fundamentals and prompt engineering, with proven experience improving model output by tuning context, persona, and examples.
  • Solid data analysis ability with the capacity to spot strategic issues quickly from metric changes.
  • Self-driven, outcome-oriented, and comfortable influencing stakeholders in fast-moving, ambiguous environments.

Why Join TikTok

TikTok aims to inspire creativity and bring joy through short-form video. The company’s global headquarters are in Los Angeles and Singapore, with additional offices in New York City, London, Dublin, Paris, Berlin, Dubai, Jakarta, Seoul, and Tokyo.

Culture and Inclusion

The company emphasizes curiosity, humility, resilience, and a constant-improvement mindset. It also values diverse perspectives and is committed to building an inclusive workplace that reflects the global communities it serves.

Leave it if you'd like a reply — we won't use it for anything else.

Click to browse, drag & drop, or paste a screenshot

PNG, JPG, GIF, MP4, WebM, MOV · Max 20MB each · Up to 5 files