Subchapter 23.5
references/decision-refs/eks.mdMarkdown1 KBView on GitHub
Kubernetes, full control, portable across clouds, GPU-capable.
Existing K8s cluster, platform-engineering team, multi-cloud portability, GPU workloads.
None that eliminate it for agents (it’s the GPU / multi-cloud / full-control winner).
Highest ops burden; only worth it with existing K8s or GPU/multi-cloud needs.
Hands off to the source-platform migration skill (gcp-to-aws or heroku-to-aws) for compute-layer config.
Entry: container behind Ingress/Service exposing your HTTP/gRPC endpoint. IAM: IRSA for pod-level permissions with bedrock:InvokeModel (model-bearing units only — a model-less unit omits it) + service-specific permissions. Networking: in-VPC by nature; Service/Ingress endpoints over TLS; VPC endpoints for AWS services.