Federated learning (FL) projects often begin with a straightforward setup: one server, a few clients, and one dataset at each site. As those projects grow, the challenge shifts from running an algorithm to operating shared infrastructure. GPUs must be allocated when jobs need them, multiple research studies must remain separated, and every participating organization must retain control of its own data, secrets, and compute policies.

That operational complexity grows when organizations rely on different environments. One site may use a Docker host, another may run a Kubernetes cluster, and a research center may schedule its GPU workloads with Slurm.…