Project Associate — AI Research & Data Engineering (AI4Bharat, IIT Madras)
Worked on RLHF-style feedback and response quality evaluation for large language model research. Assisted with preference labeling and structured labeling to support training and quality assurance workflows. Performed dataset quality checks to ensure label consistency for downstream model fine-tuning. • Conducted data collection, cleaning, and preparation for training datasets • Supported preference labeling and output quality scoring for RLHF-style pipelines • Helped validate label quality and consistency using automated checks • Collaborated with research teams to optimize labeling and evaluation processes