At Invisible Technologies, I evaluate and train frontier LLMs across reasoning, factuality, multilingual, and safety dimensions. I also contributed to Humanity’s Last Exam, a Nature-published benchmark, designing and evaluating expert-level questions about the limits of frontier AI reasoning.
My research work spans applied chemistry and materials science, including polymer formulations, biodegradable hydrogels, and composite materials. At the National Tobacco Testing Laboratory, I conducted instrumental analysis and developed tobacco protocols, and integrated Python-based frameworks for automated data processing.
