AI Data Annotator & Prompt Engineer (Freelance, Remote)
Provided human evaluation of AI-generated text across diverse domains to judge accuracy, tone, clarity, safety, and instruction-following. Applied rubrics to identify errors, bias, and hallucinations for model output quality improvement. Produced structured feedback suitable for RLHF workflows and iterative refinement. • Evaluated AI text for factual accuracy and instruction compliance • Assessed safety and appropriateness of responses • Flagged bias and hallucination risks using rubrics • Delivered actionable human feedback for model improvement