Expert-dependent procedures
The safe path lives in one engineer’s sequence of checks, commands and exceptions.
Self-hosted control Deterministic operations Reviewable evidence
Self-hosted Kubernetes operations control plane
Turn fragile Kubernetes runbooks into controlled operations.
ClusterPilot gives platform, operations and QA teams one customer-hosted path to plan, execute and prove critical Kubernetes lifecycle work—without handing operational authority to an external SaaS control plane.
Bring us the workflow that hurts most · hello@clusterpilot.de
Lifecycle state at a glanceProvider · nodes · preparation · operations
Plan before changeReview what will happen before privileged execution begins
Bound the executionUse typed work, compatible agents, timeouts and explicit gates
Retain the evidenceKeep state, logs, results and artifacts attached to the operation
Keep customer controlHost the control plane, data and execution boundary in your environment
The operational gap
Kubernetes is automated. The critical handoffs around it often are not.
Infrastructure tools provision resources. CI/CD ships applications. Monitoring reports symptoms. But high-impact cluster work still crosses tickets, scripts, terminal sessions and team boundaries—with no single state contract connecting intent, execution, recovery and proof.
Explore use cases →Risk discovered during execution
Compatibility, scope and destructive impact become visible after work has already started.
Evidence assembled afterward
Operations, QA and security rebuild the timeline from logs, screenshots and conversations.
See the actual platform
From cloud account to accepted cluster—one connected path.
Validate the provider boundary, provision role-based infrastructure and follow Kubernetes lifecycle work without losing the operational context between systems.
01 · Connect
Validate a provider account
Register write-only credentials, validate the provider boundary, and continue only when the account is ready for infrastructure.
02 · Provision
Plan role-based infrastructure
Review provider-safe defaults, node roles and the deterministic plan before infrastructure provisioning starts.
03 · Operate
Follow every operation
Track ordered lifecycle and add-on work with status, actor, duration, stable failure context and an explicit recovery path.
Why ClusterPilot
Make operational knowledge repeatable without creating a black box.
ClusterPilot keeps every high-impact boundary visible while giving teams one model they can learn, operate and review together.
Replace fragmented runbooks with one operating contract
Bring provider infrastructure, node preparation, cluster lifecycle, add-ons and validation into repeatable, reviewable workflows.
Explore this capability →Expose risk before privileged work starts
Review intended resources, immutable inputs, compatibility, authorization, feature gates and destructive intent before execution.
Explore this capability →Keep the evidence with the operation
Use durable state, ordered steps, stable errors, bounded logs, artifacts and recovery context instead of reconstructing terminal history.
Explore this capability →Before and after
Stop rebuilding the operational truth after every change.
ClusterPilot turns disconnected activity into one durable lifecycle that teams can inspect before, during and after execution.
MomentFragmented workflowWith ClusterPilot
Change requestTicket text and tribal knowledge Reviewed plan with immutable inputs
ExecutionScripts, terminals and disconnected jobs Bounded work through a compatible lifecycle agent
ProgressPartial logs and manual status checks One durable OperationRun with ordered steps
FailureReconstruct the last known state Stable error, evidence and explicit next action
HandoverScreenshots and copied command output Identified artifacts, checksums and audit context
Customer-hosted by design
Standardize operations without surrendering the control plane.
Deploy ClusterPilot’s API, workers, data integration, artifact storage and lifecycle agents in your environment. Connect providers, registries, target hosts and clusters through explicit trust and credential boundaries.
- No mandatory SaaS control plane
- Customer-owned PostgreSQL and evidence storage
- Digest-pinned API and agent images
- Public CA, private CA and restricted-network paths
Customer environment
Web / APIPostgreSQLArtifacts
ClusterPilotPolicy · Orchestration · Evidence
Lifecycle agentsPrivate registryTarget clusters
One operating language
The same lifecycle for every critical change.
Operators learn one model for cluster creation, scale, upgrade, replacement, add-ons and validation.
Plan
Show intended resources, steps, dependencies, compatibility and risk.
Gate
Evaluate identity, provider, agent, capability and approval boundaries.
Execute
Assign bounded typed work inside the customer-controlled environment.
Evidence
Retain state, events, logs, results, artifacts and correlation.
Recover
Cancel, retry, resume cleanup or restore through an explicit contract.
Start with a real workflow
A technical conversation, not a generic sales funnel.
We start with your environment, your operational pain and the evidence your teams need before expanding the scope.
Use-case review
Choose the operational workflow that creates the most risk, delay or expert dependency today.
Technical walkthrough
Map ClusterPilot’s control plane, agent, registry and evidence boundaries to your environment.
Evaluation plan
Define a bounded non-production path, acceptance criteria, responsibilities and the evidence required to proceed.
Bring us your hardest workflow
Let’s find out where ClusterPilot creates operational leverage.
Tell us about your Kubernetes estate, the lifecycle task that consumes the most expert time, and what a successful evaluation must prove.
hello@clusterpilot.de