AI infrastructure startup building the Mojo programming language and MAX runtime to let AI inference code run optimized across any chip without hardware-specific rewrites.

Builds full-stack AI networking infrastructure to connect accelerators, memory, and storage into high-performance AI data center fabric.