AI Chatbot Quality Annotation
Evaluated and annotated AI generated responses for quality, accuracy, helpfulness, and safety as part of reinforcement learning from human feedback[RLHF] pipeline. Reviewed and ranked model outputs across multiple criteria, maintaining consistency with annotation guidelines. Labeled over 500 text samples with attention to factual correctness and tone appropriateness.