AI Response Reviewer
As an AI Response Reviewer, I evaluated over 1,200 AI-generated responses weekly for reasoning accuracy, clarity, policy compliance, and instruction adherence. I used a structured scoring rubric to ensure consistent and reliable output assessments. My responsibilities included identifying reasoning flaws, flagging hallucinations, and maintaining high audit accuracy. • Evaluated outputs for factual accuracy and policy alignment. • Provided written rationales for unsafe or unsupported responses. • Increased evaluation consistency across review cycles. • Consistently met strict turnaround and quality audit targets.