AI Trainer & Data Annotator (Freelance) — Outlier, DataAnnotation.tech, Prolific
Provided prompt evaluation and response ranking to improve LLM accuracy and safety. Performed quality assurance on model outputs by checking instruction adherence and identifying response issues. Worked across freelance AI training platforms to support high-volume annotation requirements. • Conducted prompt-response ranking and evaluation • Assessed responses for factual errors and hallucinations • Verified adherence to task-specific rubrics • Maintained high inter-annotator agreement where applicable