AI Trainer / Data Annotator (Freelance or Practice) — Online platforms
Performed AI-generated response evaluations for accuracy, clarity, and relevance across multiple topics. Wrote and compared prompts with ideal/expected responses to ensure outputs followed required guidelines. Identified common issues such as bias, hallucinations, and unsafe content, then ranked outputs based on performance criteria. • Reviewed AI responses against detailed instructions • Detected quality and safety failures (bias, hallucinations, unsafe content) • Ranked and compared outputs using rubric-based criteria • Ensured consistency with provided labeling guidelines