AI Trainer / LLM Evaluator
As an AI Trainer and LLM Evaluator, I evaluated and annotated AI-generated text outputs for factual accuracy, relevance, safety, clarity, tone, and instruction-following. I performed structured side-by-side comparison, supported RLHF workflows, rated AI responses, and provided written feedback for model improvement. I also rewrote and categorized prompts and responses, enforced policy compliance, and reviewed Wikipedia-style articles. • Conducted prompt and response categorization by task, quality, and risk level. • Identified hallucinations, unsupported claims, and unsafe or low-quality outputs. • Moderated AI/user-generated text content for policy and safety compliance. • Used structured rating rubrics and quality guidelines for model output evaluation.