AI Trainer
Evaluated and rated AI-generated responses using structured feedback workflows to improve model accuracy, helpfulness, and safety. Authored and refined prompt-response pairs across technical domains to strengthen training datasets. Identified edge cases, factual errors, and policy violations to inform iterative improvements to large language model behavior. • Provided response ratings and structured feedback for RLHF pipelines. • Built prompt-response pairs for technical topics including coding, reasoning, and system design. • Checked model outputs for safety, factuality, and policy compliance. • Contributed to continuous iteration of model training and evaluation cycles.