AI Data Quality & Evaluation Analyst

TikTok; Kuala Lumpur; On-site; MYR 56,000 - 89,000; Full time

Job summary

TikTok’s AI Data Service & Operations team builds multilingual data assets for safety and non-safety domains, acting as the gatekeeper for training and evaluation data. As a graduate, you will help define quality standards and drive consistency across models and governance, with opportunities to tackle bold ideas and grow quickly.

You will work with cross-functional teams to refine judgment rules, review data outputs, and ensure trustworthy data throughout the production pipeline.

Qualifications

• Bachelor's degree in data, statistics, computer science, law, linguistics, or social sciences preferred.
• Experience in data quality, quality assurance, content moderation, compliance, content operations, editorial review, or AI data/annotation is a strong plus.
• Excellent English proficiency (listening, speaking, reading, writing) with understanding of English video/text content; IELTS 7.5 / TOEFL 105 or equivalent.
• Proficient in Excel and common data tools; hands-on experience with mainstream LLM products is preferred.
• Strong logical thinking, problem decomposition and detail orientation.

Responsibilities

• Help design and iterate on data quality standards and acceptance criteria for LLM training and evaluation scenarios.
• Review and adjudicate data outputs against quality standards and correct issues to ensure accuracy.
• Investigate edge cases and turn conclusions into reusable judgment rules and knowledge assets.
• Track quality metrics and root-cause issues; drive improvements in production processes.
• Partner with cross-functional teams to align on quality standards and feedback loops.

Skills

Data quality

Quality assurance

Content operations

AI data / annotation

English proficiency

Problem solving

Education

Bachelor's degree in data/statistics/CS/law/linguistics/social sciences

Tools

Excel

LLM platforms (ChatGPT/Claude/Gemini)

Job description

TikTok’s AI Data Service & Operations team builds multilingual data assets for safety and non-safety domains, acting as the gatekeeper for training and evaluation data. As a graduate, you will help define quality standards and drive consistency across models and governance, with opportunities to tackle bold ideas and grow quickly.

You will work with cross-functional teams to refine judgment rules, review data outputs, and ensure trustworthy data throughout the production pipeline.