Soumith Chintala

Soumith Chintala is a software engineer and AI researcher best known as the co-creator of PyTorch, the open-source deep learning framework that has become the dominant tool for AI research and much of production machine learning worldwide. Born and raised in Hyderabad, India, he earned a B.Tech in Information Technology from Vellore Institute of Technology before completing a Master's in Computer Science at New York University, where he focused on artificial intelligence and deep learning.

Chintala joined Meta (then Facebook), where he spent roughly eleven years building out its AI infrastructure. In September 2016 he led the release of PyTorch's first version alongside collaborators including Adam Paszke, Sam Gross and Gregory Chanan. PyTorch's dynamic computation graph and Python-native design made it dramatically easier to use than earlier frameworks, and it rapidly overtook rivals to become the standard tool taught in university courses and used in most published AI research.

Over the following decade, Chintala oversaw PyTorch's evolution from a research tool into critical global AI infrastructure, eventually moving it under the neutral governance of the PyTorch Foundation within the Linux Foundation in 2022 to broaden industry investment beyond Meta alone. Within Meta he rose to vice-president of AI infrastructure, a role that put him at the centre of the company's large-scale AI model training systems.

Chintala left Meta in November 2025 and joined Thinking Machines Lab, the AI startup co-founded by former OpenAI chief technology officer Mira Murati. In 2026 he was named the company's Chief Technology Officer, taking responsibility for its technical strategy, research direction and AI infrastructure scaling as the lab competes at the frontier of AI model development.

Chintala's move from the world's most widely used open-source AI framework to the technical leadership of one of the industry's most closely watched new labs reflects both his standing as an infrastructure builder and the broader talent migration reshaping frontier AI. PyTorch itself remains his most enduring legacy: the substrate on which most of today's large language models and AI research are built.