Data Annotator / Data Labeling (Atlas capture)
Evaluated pair-wise search engine results to select the better option and improve recommendation performance. Compared AI responses to prompts to ensure safety, helpfulness, and factuality. Reviewed annotator submissions and provided constructive feedback to improve quality and consistency. • Performed relevance comparison between paired outputs • Checked responses against safety and factuality criteria • Audited other annotators’ work • Delivered feedback aimed at reducing errors