Senior AI Trainer & RLHF Specialist at Public AI (Remote)
Led RLHF initiatives to fine-tune generative language models using structured human feedback to improve reasoning and reduce hallucinations. Reviewed and ranked AI-generated responses using accuracy, tone, and safety guidelines while supporting reward-model development. Developed comprehensive annotation guidelines for large-scale labeling with a focus on high inter-annotator agreement and data quality. • Perform response evaluation and ranking for training signals • Create annotation guidelines and maintain labeling consistency • Collaborate with ML engineers on edge cases and dataset iteration • Use prompt engineering to test model performance across domains