Ryan Lagasse

WORK HISTORY

EDUCATION

ABOUT RYAN LAGASSE

Ryan Lagasse is an AI/ML Research Engineer at Lockheed Martin, where he focuses on interpretability and reasoning systems within large-scale defense AI projects. His work centers on understanding why large language models behave the way they do - moving beyond performance metrics toward genuine alignment and mechanical comprehension.Ryan leads research in model interpretability, alignment, and control schemas, with an emphasis on making AI systems transparent and predictable rather than opaque black boxes. His experience spans developing transformer-based architectures, alignment research at UConn’s Advanced Computing Center, and mentoring research teams at Algoverse whose work has been accepted at top-tier conferences such as ICML, NeurIPS, and EMNLP.He aims to bridge the gap between empirical AI success and theoretical understanding - advancing the field toward explainable, controllable intelligence that firms can trust to reason safely and effectively.