interconnects, transit and routing design, DNS, private service endpoints, and hybrid identity and federation models. Experience architecting GPU compute for AI, including cluster orchestration (Kubernetes, Slurm, or equivalent), scheduling and multi-tenancy, accelerator selection, and high-performance interconnect (e.g., NVIDIA GPUs, CUDA, NCCL, InfiniBand/RoCE). Solid command … caching strategies, and the residency and sovereignty constraints that shape them. Sufficient hands-on fluency with infrastructure-as-code and automation (e.g., Terraform, Ansible, Kubernetes tooling) and working proficiency in Python to prototype, validate, and stress-test your own designs. Track record of building capacity and total-cost models that ...