Callozum
Inference that knows where to run
A router for AI work. It splits each request or stream into parts and decides where each part runs: on the device, on a local GPU, at the edge or in the cloud. A part that needs no model gets dropped, and your data rules decide what may leave. Built for robots, wearables, cameras and anything with rules about where data may go.
The waitlist is open. The first stream demo is in progress. · my product · 2026
- links
- live ↗
- thinking
- Inference That Knows Where to Run
Cloudflare WorkersDurable ObjectsWorkers AIClef