Data Annotator / Labeler (Freelance or Academic Project) - Self-Employed
Evaluated and labeled datasets (text, image, or audio) according to precise guidelines to improve model accuracy and performance. Conducted Reinforcement Learning from Human Feedback (RLHF) tasks by ranking and rewriting AI responses for tone, safety, and accuracy. Identified edge cases, errors, and inconsistencies in training datasets, reducing labeling error rates by maintaining high attention to detail. Consistently achieved target throughput metrics while strictly maintaining an accuracy rate above 95%.