Cartesia

Karan Goel holds a PhD in Computer Science from Stanford, a Master's in Machine Learning from Carnegie Mellon, and dual degrees in Electrical Engineering from IIT Delhi. He was a Research Scientist at Snorkel AI and a Siebel Scholar before founding Cartesia in 2023 alongside fellow Stanford AI Lab researchers. Goel and his co-founders invented State Space Models (SSMs), a sequence-modeling architecture that rivals Transformers in quality while running at a fraction of the compute cost — the technical foundation Cartesia commercializes for real-time voice and multimodal AI applications.

Albert Gu earned his PhD at the Stanford AI Lab and is an Assistant Professor of Machine Learning at Carnegie Mellon University. He co-authored the landmark Mamba paper (December 2023) with Tri Dao, introducing selective state space models that achieve five times higher inference throughput than Transformers with linear sequence-length scaling, reaching state-of-the-art performance in language, audio, and genomics. Gu co-founded Cartesia in 2023 to commercialize SSM-based foundation models, serving as Chief Scientist while maintaining his academic appointment at CMU.

Arjun Desai studied Biomedical Engineering and Computer Science at Duke before pursuing an Electrical Engineering PhD at the Stanford AI Lab, where he worked on State Space Models alongside Karan Goel and Albert Gu. He interned at Microsoft, Verily Life Sciences, and Apple's Machine Intelligence team before co-founding Cartesia in 2023. His background bridges medical AI and production ML systems, and he has published research on medical image processing and neonatal prematurity classification.

Brandon Yang holds a BS in Computer Science from Stanford and was pursuing a PhD there — advised by Christopher Ré — when the Cartesia founding team spun out in 2023. During his doctoral work he contributed to ML training and evaluation infrastructure at Snorkel AI. At Cartesia, Yang focuses on model training systems and the infrastructure that enables the company's SSM-based models to run in real time across diverse hardware environments.

Chris Ré is a Professor of Computer Science at Stanford and a 2015 MacArthur Fellow. His research spans machine learning systems, database theory, and foundation models. He has co-founded four companies from his lab's research: Lattice (acquired by Apple in 2017), Inductiv (acquired by Apple in 2020), SambaNova Systems, and Snorkel AI. He advised the PhD students who invented State Space Models at the Stanford AI Lab and joined as a co-founder when the team spun out Cartesia in 2023, providing scientific credibility and entrepreneurial experience across the ML infrastructure stack.