We're still building this site. Until it's ready, two doors are open.
Perception and reasoning models are now good enough to run robots and other machines that act in the real world. Running them there, inside a power budget and a latency budget, is still the hard part.
We're building the inference layer those systems run on: the piece between a trained model and the machine that has to act on it. We're looking for a few teams building perception and reasoning pipelines for physical systems to build it with. You get early access and a direct line to the people writing the code. We get your hardest workloads.
Physical AI is where models meet consequences. The companies that win it will need inference that is fast, predictable, and cheap enough to run everywhere a machine does. That layer doesn't exist yet.
Misliya is building it. We're early, and we're talking to investors who back infrastructure before it's obvious. If that's you, we'd like to walk you through what we're seeing and where we're headed.