Services
Consulting and training
We take on paid work on GPU kernels, MLIR, cross compilation, and getting models onto embedded targets.
Start a conversation
Bring us the hard deployment problem.
Tell us the target hardware, the model, and the constraint you are stuck on.