provisioning: azure · gcp · aws · on-prem

Your code. Your cloud account.
One click and it's provisioned.

KubeHype connects to your GitHub repo and your cloud account — Azure, Google Cloud, AWS, or on-premises — and auto-provisions the cluster, networking, and delivery pipeline behind it. Nothing leaves your account, nothing changes about who owns your compliance boundary.

khctl — one-click provision

$ khctl connect github.com/acme/checkout-api

Target: your-aws-account region: eu-north-1

Auth: federated identity — no keys stored, ever

Cluster, network, ingress, identity provisioned

Pipeline connected — push to deploy

Knative service checkout-api live, scale-to-zero on

You're all set.

You already have the code and the cloud account. That should be enough.

KubeHype isn't tied to one cloud. We bring the same one-click provisioning to Azure, Google Cloud, AWS, or your own on-premises Kubernetes — and we deploy it inside your account, under your existing controls. Nothing runs in a shared environment we own; your compliance posture doesn't change because we showed up. Commit to your GitHub project, and we provision the infrastructure and wire it straight into your pipeline.

how a commit becomes a running app

Connect once. Every push after that just deploys.

You keep your code and your cloud account. We provision the path between them and keep it in sync — on Azure, Google Cloud, AWS, or your own bare-metal Kubernetes.

01 Connect Point us at your GitHub project and your cloud account. One click, federated identity — no keys handed over.
02 Auto-provision Cluster, networking, identity, and ingress stand themselves up inside your account, from reviewable IaC.
03 Pipeline wired in Your existing CI is connected to the new infra — no infra team ticket required.
04 Sync engine A GitOps reconciliation loop keeps the cluster matching what's committed, and promotes across stages on policy.
05 Running app Deployed as a Knative service where it fits — scale-to-zero, autoscaled, behind private networking.
what one click actually provisions

The auto-provision catalog

Pick a target cloud — or none, if you're on-prem. Every item below stands up in your account, wired together, ready for your pipeline to push to.

cluster

Kubernetes control plane

AKS, GKE, EKS, or a self-managed cluster on your own hardware — provisioned to the same hardened baseline regardless of where it runs.

serverless

Knative runtime

Installed and configured out of the box. Your services deploy as Knative apps with scale-to-zero and request-based autoscaling, no YAML required.

network

Private networking

Private endpoints and egress rules provisioned per cloud's native model — no service is reachable from the public internet unless you say so.

identity

Federated workload identity

Pipelines and pods authenticate via short-lived, federated tokens. No cloud keys are ever generated, stored, or handed to us.

delivery

GitOps sync & promotion

A reconciliation loop keeps your cluster matching Git, plus a promotion pipeline that moves builds dev → staging → prod on rules you set.

ingress

Ingress & TLS

Managed certificates and routing provisioned per service, renewed automatically — never a manually issued cert again.

secrets

Secrets management

Wired to your cloud's native secret store — Key Vault, Secret Manager, or Secrets Manager — so nothing sensitive ever lives in a manifest.

observability

Metrics, logs & alerts

A baseline observability stack ships with every cluster — dashboards and alerting from day one, not something bolted on later.

finops

Cost attribution

Spend broken down per namespace and service from the start, so cost is visible before it's a surprise.

bring your own cloud

We provision inside your account. We don't become your account.

No shared tenancy, no new vendor sitting between you and your cloud bill. Your existing controls — IAM, audit logs, compliance scope — stay exactly where they are.

azure

Azure

AKS, Private Link, and Managed Identity — provisioned across as many subscriptions as you run, from Terraform you can read.

gcp

Google Cloud

GKE, Private Service Connect, and Workload Identity Federation — same one-click path, native to how GCP does identity and networking.

aws

AWS

EKS, PrivateLink, and IAM Roles for Service Accounts — provisioned per account, per region, no cross-account access requested.

on-prem

On-premises

Your own hardware, your own data center. We provision the same GitOps and delivery layer on top of a self-managed cluster.

compliance

Your compliance boundary, untouched

We never hold standing access. Provisioning runs through federated, time-boxed identity scoped to exactly what it needs to build — then it's gone.

engagement models

Four ways to bring us in

Start with the platform, or start with the parts that hurt most today.

01One-click provisioning, in your account
+

Connect your GitHub project and your cloud account — Azure, Google Cloud, AWS, or on-prem — and we provision the cluster, network, and identity footprint directly inside it. You keep ownership and the bill; we keep it running.

  • Reviewable IaC, not a black-box wizard
  • Federated, time-boxed access — no standing credentials
  • Same one-click path regardless of target cloud
02Managed delivery pipeline
+

Your Helm charts and Knative services, reviewed, versioned, and promoted automatically instead of hand-applied. Every change traces back to a commit.

  • Chart structure and values-file conventions audited
  • Automatic promotion dev → staging → prod on rules you set
  • Rollback is git revert, not an incident
03Cloud FinOps
+

Per-service cost visibility across whichever cloud you're on, plus concrete right-sizing recommendations — not a generic savings report.

  • Namespace-level cost attribution on shared clusters
  • Nodepool and reserved-capacity right-sizing
  • Monthly variance review against forecast
04Migration & consultancy
+

Moving from self-managed clusters, per-team sprawl, or another cloud entirely — planned and executed with rollback checkpoints at every stage.

  • Current-state assessment: identity, networking, secret handling
  • Staged cutover plan with defined rollback points
  • Knowledge transfer so your team owns it after we leave
get in touch

Talk to the team running this

Tell us about your current setup — subscription count, cluster topology, what's actually breaking — and we'll respond with specifics, not a sales deck.

Email directly

For a quick question, or if you'd rather skip the form.

Send us your setup

Fill this in and it opens a prefilled email in your client — nothing is sent from this page.