Parallel Works
ACTIVATE Kubernetes

Any Kubernetes Cluster, One Control Plane

Bring the Kubernetes clusters you already run under one identity, one budget, and one operations view — without a migration project.

Bring Your Own Clusters

Attach existing Kubernetes in minutes

Point ACTIVATE at your kubeconfig and the cluster shows up as a compute provider alongside AWS, Azure, GCP, OCI, and OpenStack. No rebuilds, no migration project — your existing workloads, operators, and GitOps tooling keep running untouched.

  • Works with EKS, GKE, AKS, OpenShift, Rancher, and upstream Kubernetes
  • Attach in minutes from the ACTIVATE UI; detach cleanly any time
  • Coexists with ArgoCD, Flux, and your existing operators
  • Manage many clusters from one organization
ACTIVATE • Bring Your Own Clusters

Namespaces You Already Have, Mapped to Budgets

Tag a namespace, cost follows

ACTIVATE works with the namespaces you already operate. Map any namespace to an allocation and its cost rolls up to that allocation — alongside HPC jobs, AI tokens, and user sessions. Finance gets per-team, per-project chargeback across every cluster you connect, in one dashboard.

  • Use the namespaces you already have — no forced layout, no CRDs
  • A single namespace annotation drives chargeback mapping
  • One allocation can span many namespaces and many clusters
  • Rolls up with HPC and AI in one finance view — no spreadsheets
ACTIVATE • Namespaces You Already Have, Mapped to Budgets

Authorization Stays in Your Cluster

Your RoleBindings, your policy

ACTIVATE acts as the OIDC issuer for connected clusters and signs short-lived tokens with your users' group membership — but every authorization decision is made by your kube-apiserver against the RoleBindings you already manage. No parallel policy engine, no shadow permissions.

  • Native Kubernetes RBAC — RoleBindings stay under your control
  • Short-lived OIDC tokens minted by ACTIVATE, signed with platform keys
  • Group claims (parallelworks:-prefixed) map to your existing RoleBindings
  • No long-lived service-account tokens handed to end users
ACTIVATE • Authorization Stays in Your Cluster

Day-2 Ops Without Leaving the Browser

Workloads, logs, and pod exec — one tab

Browse every workload across every connected cluster, inspect pod status, tail logs, and exec into a container — all under the same SSO and RBAC. Less kubectl context-switching for the developers shipping, less screen-sharing for the platform team supporting them.

  • Unified workload list across clusters, namespaces, and types
  • Pod detail, container status, and live log streaming
  • In-browser exec shell into any pod the user is authorized for
  • No new tools to install — same identity as the rest of ACTIVATE
ACTIVATE • Day-2 Ops Without Leaving the Browser

Built for Platform Teams

Bring an existing Kubernetes cluster under ACTIVATE governance without rebuilding it

No Lock-in

Your cluster stays yours; ACTIVATE attaches and detaches cleanly

Multi-Cluster

Manage many clusters from one ACTIVATE organization

GPU-Aware

Attribute and chargeback GPU usage per namespace

GitOps-Friendly

Coexists with ArgoCD, Flux, and your existing operators

Bring Your Cluster Into ACTIVATE

Connect an existing Kubernetes cluster and inherit unified identity, allocations, and chargeback in one afternoon.