Content Quality and Evaluation Specialist (Korean Speaking) - Safety Model Operations

Written by the T&S Network from TikTok's posting.

Builds and maintains golden sets, the ground-truth labeled data used to evaluate TikTok's automated content safety models, along with the quality assurance around human review for those models. Sits in the Safety Model Operations team, whose delivery group reviews labeling accuracy and works to improve AI model performance across regions.

What you'd do

• Build and iterate golden sets of long-tail, high-risk and edge-case content as ground truth for policy, model evaluation and enforcement teams.
• Run deep-dive analyses on high-risk and heavily debated cases to find misapplication patterns and risk trends.
• Turn macro policies into executable SOPs and operational guidelines.
• Work with Operations, Algorithm and Product teams to find weaknesses from golden set metrics and improve human review and AI interception.
• Build daily quality assurance mechanisms and a content quality management framework.
• Track and attribute data metrics to raise labeling accuracy and operational efficiency.

What it asks for

• Proven experience in internet Trust & Safety, QA, AI data evaluation or policy operations.
• Strong logical reasoning and root-cause analysis in ambiguous business scenarios.
• Professional working proficiency in English, with strong written and verbal communication.
• Korean proficiency for daily liaison with the Korea local team.
• Understanding of how machine learning models work and how labeled datasets are structured (preferred).

Our best summary of the posting, and it can be wrong. The employer's posting is what counts: read it in full on TikToknorth_east before you apply. If we got something wrong, report this listing.