Skip to content

Overview

Kupe Cloud is a managed Kubernetes platform built around isolated tenant clusters with shared platform services. Your team works inside standard Kubernetes clusters and APIs, while Kupe operates the surrounding control plane, networking, observability, access, and cluster lifecycle.

The result is a platform that feels like Kubernetes, but removes the need to build and run the supporting stacks yourself.

Each managed cluster gives you an isolated Kubernetes environment for your workloads. Kupe then layers shared platform capabilities around those clusters:

  • Cluster lifecycle through the console, API, and Terraform
  • GitOps delivery through centrally managed Argo CD
  • Access and authentication through SSO-backed console and kubeconfig flows
  • Networking through shared Gateway API infrastructure, DNS, and TLS
  • Observability through managed Grafana, metrics, logs, alerts, and notifications
  • Secrets through tenant-isolated vault-backed secret storage and sync

Your team still deploys standard Kubernetes resources. The platform is opinionated about how clusters are operated, not about the applications you run inside them.

Kupe operates the underlying platform and control plane so teams can focus on workloads and delivery rather than building platform plumbing.

AreaWhat Kupe manages
Cluster provisioningCreates clusters, registers them in platform systems, and makes them available in the console and CLI
Kubernetes lifecycleSupported versions, upgrade workflows, health checks, and platform-side coordination
InfrastructureHost nodes, networking, storage integration, ingress infrastructure, and platform service deployment
Authentication and accessSSO integration, tenant roles, kubeconfig generation, and access boundaries
GitOps control planeShared Argo CD, tenant projects, and destination registration for managed clusters
Observability stackGrafana, Loki, Mimir, Alertmanager, and baseline dashboards
Secret storage and syncTenant-isolated vault and cluster sync workflows

Your team owns the workloads and operating practices inside your clusters.

AreaWhat your team owns
Application manifestsHelm charts, Kustomize overlays, or plain YAML stored in Git
Namespace designHow workloads are grouped, named, and separated inside a cluster
Delivery workflowPromotion rules, repository layout, sync policies, and rollback practices
Runtime configurationEnvironment variables, secrets usage, routing rules, and dependency configuration
Service healthSLOs, dashboards, alert rules, incident response, and capacity decisions

The normal flow looks like this:

  1. Create a cluster in the console, API, or Terraform.
  2. Download kubeconfig or use the console for day-to-day access.
  3. Register or connect a Git repository in Argo CD.
  4. Deploy workloads into the cluster through GitOps.
  5. Expose services with HTTPRoute.
  6. Monitor health in Grafana, logs, and alerts.
  7. Manage secrets, upgrades, and day-2 changes through the platform.

This keeps the operating model consistent across teams and clusters, which is one of the main reasons to use the platform in the first place.

You can work with your clusters through several interfaces, depending on the task:

InterfaceBest for
ConsoleCluster creation, inspection, workload troubleshooting, secret management, and operational tasks
CLI kubeconfigkubectl, Helm, scripts, and local engineering workflows
Argo CDGitOps application deployment and sync status
APIProgrammatic cluster, member, secret, and API key management
TerraformDeclarative provisioning and platform changes as code

Use the console and CLI for day-to-day cluster operations. Use the API and Terraform when you want repeatable automation around the same platform objects.

  • Concepts — the core platform ideas and how the main pieces fit together
  • CLI Access — interactive and automation kubeconfig flows
  • Secrets — store values once and sync them into one or more clusters
  • Limits — practical bounds for clusters, secrets, and resource usage