United StatesGuy Gur-Ari
Trained as a theoretical physicist (PhD, Weizmann Institute, 2014; postdoctoral research at the Institute for Advanced Study in Princeton and Stanford), Gur-Ari spent nearly a decade at Google as a Senior Staff Research Scientist. There he led a team focused on understanding and improving large language models, contributing directly to the development of the PaLM models (precursors to Gemini). That work — particularly on how LLMs reason about structure and context at scale — directly informs Augment's architecture. His public posture is defined by a near-obsessive focus on latency: he has described building a coding assistant that returns results in ~200 milliseconds as a core design constraint, arguing that speed is what makes AI assistance usable in real developer workflows.
