Artificial intelligence that breaks free from the datacenter

Our mathematically-pioneering approach, developed at Caltech, solves a decades-old research problem: how to dramatically compress neural networks without sacrificing their reasoning power.
We produce models that are 16x smaller, 8x faster, and 4x more energy efficient, running on phones, laptops, and edge devices instead of billion-dollar infrastructure.

Our mission

We believe AI's future won't be defined by who builds the largest datacenters, but by who most efficiently converts energy into intelligence. Just as computing shifted from massive supercomputers to devices in everyone's hands, AI will follow the same path.
Our models keep sensitive data on-device, enabling privacy-first applications in healthcare, defense, and finance that were previously impossible at the edge.
This isn’t just theory. Our flagship 1-bit Bonsai 8B model proves what's possible: 8 billion parameters compressed to a 1GB footprint, delivering performance competitive with full-precision models at a fraction of the cost. It's the world's first 1-bit LLM designed for real-world deployment anywhere: from phones to laptops to robotics.
Bonsai 8B is just the beginning. Our techniques scale to larger models, offering industry-leading efficiency gains for cloud providers, enterprises, and governments managing energy and operating costs.

Meet the team

Babak Hassibi

Co-Founder & CEO

Sahin Lale

Co-Founder, Head of Research

Omead Pooladzandi

Head of Innovation & Research

Reza Sadri

Co-Founder & VP, Strategy

Shayan Ilbagian

CFO

Karim Mattar

VP, Engineering

Tushar Bansal

Dir, Product

Ion Stoica

Advisor

Buno Pati

Advisor

Julie Schoenfeld

Advisor

Join the team

Join us to turn mathematically rigorous research into models that run anywhere. If you’re excited to push the frontier of efficient intelligence and ship it into the real world, we’d love to work with you.

Staff AI/ML Engineer – Edge & Consumer AI

Staff AI/ML Engineer – Large-Scale & Low-Precision AI