AI Language Evaluator / LLM Data Annotator — RWS (Remote)
Evaluated AI-generated responses in Brazilian Portuguese for fluency, clarity, and contextual accuracy following detailed guidelines. Analyzed grammatical, syntactic, and semantic aspects, focusing on natural language flow and error identification. Applied structured annotation to classify model outputs and provide feedback for dataset improvement. • Evaluated responses for fluency and instruction alignment • Identified grammatical and contextual inconsistencies • Maintained annotation quality for LLM training • Reviewed large-scale dataset outputs for training purposes