Calibrate AI translation quality by manually evaluating Korean-language model outputs against the original Chinese source content, including pass/fail decisions and error severity classification, and comparing translation performance across different models (e.g. Grok, Claude).
Build benchmark datasets by selecting and annotating high-quality Korean translation samples for model evaluation and future iteration.
Validate historical annotations by reviewing previously flagged Korean translation errors and annotation rationale against the relevant Chinese source content, assessing whether the original judgments were accurate, and correcting mislabels or inconsistencies.
Develop and maintain the Style Guide and translation rules for the Korean-language based on observed recurring error patterns, translating linguistic findings into actionable language rules, terminology standards, and translation workflows.
Contribute to language-related process design discussions through working with product and development teams to improve annotation standards and rule prioritization, including how Style Guides, Glossaries, and Translation Memory (TM) should be applied.
Support blind-test evaluations by independently scoring outputs from new or updated models (e.g. Grok 4.5) in blind tests and provide structured feedback on Korean translation quality and error patterns.
Ensure game-specific localization accuracy by applying each game’s lore, character relationships, terminology, character voice, and tone conventions consistently when evaluating translations, labeling samples, and developing language rules.

Linguistic Specialist (Korean) (Contract)