Modular builds AI infrastructure software — the Mojo programming language and MAX AI compute platform — designed to unify and speed up AI model deployment across hardware and break Nvidia's CUDA lock-in.

GPU-as-a-service cloud provider delivering NVIDIA-based compute clusters for AI training and inference to underserved mid-market and enterprise customers, pursuing a Nasdaq direct listing under ticker QMLS.