AI Training Specialist / AI Model Evaluation Contributor (Mercor AI)
Trained and evaluated AI systems by reviewing model responses, comparing outputs, and selecting the most accurate and policy-aligned answers. Created and refined prompts, rating rubrics, and written feedback to improve model reasoning, factual accuracy, tone, and instruction-following. Performed quality review work including response ranking, fact-checking, and identifying issues such as hallucinations, weak explanations, unsafe responses, or incomplete answers. • Reviewed and compared AI outputs across technical, research, and general knowledge tasks • Built prompt/rubric artifacts to guide evaluation • Produced written feedback for model improvement • Conducted fact-checking and hallucination detection to support safe responses