I’m standing up a brand-new, production-grade Kubernetes environment that must run both on-premise and in the cloud. The on-premise footprint will go live first, then extend to at least one public provider, so every design choice has to support a true hybrid topology from day one. My biggest priorities are multi-tenancy and cost accountability. I need hierarchical namespaces and strict resource governance implemented side-by-side, along with a full chargeback pipeline built on Kubecost, OpenCost or a comparable Prometheus aggregation you are comfortable with. Everything has to be reproducible through Infrastructure as Code—Terraform is preferred, though CloudFormation is acceptable for the cloud portion. You should have at least five years in DevOps/SRE or infrastructure architecture and three or more that are hands-on with Kubernetes itself. I’ll rely on you to: • Design the cluster architecture, networking, storage and security controls • Build the clusters from scratch via IaC, validating that they meet production SLAs • Configure multi-tenancy (hierarchical namespaces + resource quotas/limits) with documented guardrails • Wire up cost visibility dashboards and per-tenant chargeback reports • Provide automation scripts in Python or Bash and solid Linux server hardening • Hand over clear runbooks, diagrams and a CI/CD pipeline so future upgrades are friction-free Acceptance criteria: 1. `terraform apply` (or CloudFormation stack) brings up identical clusters in test and prod. 2. Namespace hierarchy enforces CPU/memory limits and is verified with load tests. 3. Cost reports show per-tenant spend within 5% of Prometheus raw metrics. 4. All documentation is version-controlled and passes a peer review walk-through. If this sounds like the sort of challenge you enjoy and you have a track record of delivering reliable, secure hybrid Kubernetes platforms, let’s talk about timelines and the first milestone.