Own operational excellence, resilience, and platform delivery: define SLOs, on-call and incident management practices, and lead post-incident reviews; own backup, disaster recovery, and cross-environment replication for high availability (Velero, CloudNativePG, Cilium ClusterMesh), patching, security hardening, and supply-chain hygiene; and partner with DevOps Engineers on CI/CD/CV (ArgoCD) and infrastructure-as-code / GitOps (Helm/OpenTofu) so environments are reproducible and drift-free. Own GPU capacity, platform economics, and vendor selection: forecast demand, set quotas and cost attribution, and manage the infrastructure budget; evaluate commercial against open-source options, run proofs of concept, manage strategic vendor relationships, and work with procurement on long-lead hardware and licensing, making the commercial case for capacity to senior leadership.