Computer vision is very established but is very difficult to deploy and manage especially for non-specialized teams. This project helps make vision workloads reusable and composable. The goal is to make it much easier to manage vision workloads at scale in complex environments.
