The typical client
An SME from Germany, Austria or Switzerland with its own software or services: a software company with 5 to 50 developers, a manufacturer whose product has a cloud component, or an IT department that runs individual servers today. Or a company that already has a cluster but no reliable monitoring and no orderly troubleshooting.
What you have after four weeks
A Kubernetes cluster on Talos Linux, an immutable operating system without SSH, with its configuration entirely in Git. GitOps with Argo CD: every change goes through a pull request, every drift is flagged. Metrics and logs for all applications, alerts with a defined meaning in your chat, external availability checks and service level objectives with error budgets. Daily backups with one rehearsed restore. And an agent with runbooks for the most common alerts, an operations manual and training for your team.
How the agent helps with troubleshooting
When an alert fires, your team asks the agent in the chat. It examines metrics, logs and events in the cluster, follows the alert's runbook and presents a diagnosis with its reasoning and a proposal. It has read-only access to the cluster and cannot change anything itself. The fix always goes through Git: a person approves the pull request, Argo CD rolls it out, and the alert resolves. You choose the agent's language model.
How we run our own platform
Our platform runs exactly this way: 9 nodes on Talos Linux with Kubernetes 1.35, 107 applications rolled out from Git by Argo CD, 112 service level objectives with error budgets and external availability checks for more than 30 services. Our troubleshooting agents have read-only access to the cluster. As of 31 October 2026.
Where the cluster runs
On your own hardware, with a provider of your choice in Switzerland or the EU, or on our platform in the EU. The configuration belongs to you and lives in your Git repository, so you can take over operations yourself or switch at any time.
Four weeks, one rehearsed emergency
Week 1: assessment
Applications, data and today's operations, choice of location, target picture and plan. On Friday you decide: continue, or stop at a quarter of the price.
Week 2: cluster and GitOps
Building the cluster or bringing the existing one under Git, Argo CD, first application in operation.
Week 3: monitoring and backup
Metrics, logs, alerts, service level objectives, availability checks, daily backup, further applications.
Week 4: agent and handover
Read-only agent with runbooks, a rehearsed restore, a test alert your team resolves with the agent, training and operations manual.
Frequently asked questions
We already have a cluster. Does the offer still fit?
Yes. We bring the existing cluster under Git and add monitoring, backup and the agent. Whatever we find goes into the list of open items.
Can the agent break anything?
No. It has read-only access to the cluster. Every change is a pull request that a person approves.
Who runs the cluster afterwards?
You decide. Your team takes over with the operations manual and training, or we run it with on-call support for a fixed monthly price.
Do we have to rebuild our applications?
No, as long as they run as containers. Anything that does not yet goes into the list of open items.