Sahin Lale is Co-Founder and Co-Head of Research at PrismML, a Caltech spinout pioneering the world's first commercially viable 1-bit large language models for edge and cloud AI deployment.

Sahin earned his PhD in Reinforcement Learning, Control, and Machine Learning from Caltech (2016-2023), and previously held a Bachelor's degree in Electrical and Electronics Engineering from Bogazici University.

With deep expertise in neural network compression, trajectory optimization, and machine learning, Sahin co-developed the mathematical theory behind training AI models natively at 1-bit precision without sacrificing reasoning performance.

Prior to PrismML, Sahin served as Senior Staff Engineer at Neural Propulsion Systems, Inc. (2023-2025), and spent years as a Graduate Research Assistant at Caltech, where PrismML's foundational IP was developed.

PrismML's Bonsai model family delivers AI models up to 14x smaller and 8x faster than full-precision equivalents, backed by $16.25M in seed funding from Khosla Ventures and Cerberus Ventures.