Teams at WeLocalize

Annotate and manage training/test datasets, run model training, evaluate model performance, perform manual QA and error analysis, and refine annotation guidelines for NLP/ML projects.

Proofread and edit localized content for grammar, syntax, punctuation, consistency, and formatting issues. Translate occasional English content, test web applications and user interfaces on Mac OS X and iOS, document localization bugs, and verify fixes. Execute test cases, clearly communicate reproduction steps, prioritize issues, and collaborate independently or with diverse teams in a fast-paced environment, including occasional non-business hours, weekends, and holidays.

Updates machine learning training and test datasets, refines annotation guidelines, initiates model training, and evaluates model performance. The role annotates synthetic text and image data, conducts manual quality analysis, identifies error patterns, reports anomalies, and produces findings on utterance quality, LLM evaluation, ASR bugs, and customer pain points. It also applies basic NLP and linguistic knowledge to ensure data accuracy and consistency.