Sheikh Jubair

Jubair is a Machine Learning Engineer at RAD Intel, specializing in the post-training and evaluation of Large Language Models (LLMs). His research and development experience spans curriculum learning for improving domain-specific capabilities, reinforcement learning with verifiable rewards (RLVR), data-mixture optimization, and the design of domain-specific evaluation datasets. He has also developed novel approaches for improving and systematically evaluating LLM performance across targeted domains and capabilities. Previously, Jubair worked as a Senior Principal AI Researcher at HUMAIN and as a Senior Machine Learning Scientist at the Alberta Machine Intelligence Institute (Amii). He has published research in leading journals and conferences and holds a Ph.D. in Computer Science from the University of Manitoba.