AI Trainer & Data Labeling Specialist (RLHF Model Evaluation)
Served as an AI Trainer and data labeling specialist performing RLHF model evaluation for frontier AI systems. Conducted multimodal labeling and object annotation across complex image catalogs and video assets while reviewing prompts and transcripts to identify hallucinations and grade safety and logic metrics. Maintained alignment with evolving taxonomy guidelines to ensure high quality-assurance outcomes. • Multimodal data labeling and object annotation (images and video) • Prompt and response review to verify factual, logical consistency and safety metrics • Fact-checking and logic auditing against shifting taxonomy rules • Quality assurance focused on rubric adherence and error identification