AI Trainer / Multilingual Evaluator (OpenTrain AI fit / freelancer)
Performed multilingual LLM response evaluation and quality checks by comparing prompts, model outputs, and task instructions. Conducted reasoning QA and safety/instruction-following validation to ensure responses meet required criteria. Reviewed structured data alongside generated text to support accurate evaluation and feedback. • Compared prompt-output pairs to assess correctness and adherence to instructions. • Applied multilingual quality review across Chinese, Japanese, and English outputs. • Performed preference-style evaluation and reasoning QA for model behavior. • Completed structured data review and safety/instruction-following checks.