AI Data Contributor — Evaluation & Annotation (Freelance, Remote)
Provided AI response evaluation and annotation for factual, creative, and conversational prompt tasks using structured rubrics and scoring criteria. Identified hallucinations, factual inaccuracies, reasoning gaps, and unsafe outputs, documenting each issue with written rationale. Ranked multiple model outputs by helpfulness, coherence, and adherence to user intent while maintaining consistent quality across high-volume work. • Evaluated accuracy, coherence, and potential harm in AI-generated text • Applied rubric-based quality assessment and comparative response ranking • Performed instruction-following and guideline violation checks • Produced written feedback and annotations per flagged criterion