AI training and evaluation (prompt evaluation, fact-checking, hallucination auditing, and quality/rating assessments for RLHF-style optimization)
TextPerformed prompt evaluation, fact-checking, and hallucination auditing to assess the quality and correctness of AI outputs against strict guidelines. Conducted structured quality evaluations by reviewing complex information and flagging inconsistencies or unsupported claims. Delivered accurate, high-quality labeled judgments intended for AI model optimization and RLHF-style training. • Reviewed model responses for factual accuracy and guideline compliance • Audited outputs for potential hallucinations and misinformation • Applied consistent evaluation criteria for quality rating/assessment • Produced reliable evaluation labels to support training and fine-tuning workflows
Present