AI Training & Response Evaluator (Independent Freelancer, Remote)
Evaluated AI-generated responses across factual, creative, coding, and customer-service domains using detailed rubrics and quality criteria. Produced structured criterion-based written feedback describing accuracy, coherence, helpfulness, and safety issues, including failure modes and potential biases. Maintained consistent scoring across high-volume batches and flagged edge cases near policy boundaries to support quality assurance for multiple clients simultaneously. • Rated outputs for logical consistency, factual correctness, tone/coherence, and usefulness • Wrote actionable improvement reports and identified bias/failure modes for model updates • Applied rubric-based scoring standards to reduce evaluator fatigue and variability • Collaborated asynchronously with platform coordinators while meeting submission deadlines