Skip to main content
Get started

Kubernetes. Less guesswork.

An AI assistant that works with your live cluster. Investigate issues, understand the output and keep every step in context.

Try the demo Connect your cluster

Explore the demo without connecting a cluster.

Your question. Your cluster. The evidence in between.

production-eu Example cluster

Why does my API pod keep restarting?

service / api namespace: production api-4b2d api-7c9f api-8a1e Running CrashLoopBackOff Running
Ranching.Farm
The pod is hitting its memory limit.

The last container exited with OOMKilled. Compare its memory use with the configured limit before changing the deployment.

Command history kubectl describe pod api-7c9f -n production
Ask your own question in the demo

Built around the tools you already use.

kubectl Helm Kustomize

Start with the thing that isn't working.

A restarting pod. An unfamiliar namespace. A growing resource bill. Turn the symptom into the next useful check.

Find the cause. Then the fix.

Bring events, logs and recent changes into one investigation.

Kubernetes troubleshooting

Get that pod running.

Work through CrashLoopBackOff, Pending and failed health checks.

Pod debugging

See how it all connects.

Understand workloads, namespaces and the resources behind them.

Cluster management

Put idle capacity to work.

Compare usage with requests and limits before changing capacity.

Resource optimization

Bring the cluster into the conversation.

Set up your organization, connect the agent and ask your first question. Ranching.Farm brings command results back into the chat.

Set up your first cluster
  1. 1

    Add your cluster

    Give it a name in your organization. Keep your team's clusters and roles in one place.

  2. 2

    Install the agent

    Use the manifest generated for your cluster. The agent connects it to Ranching.Farm.

  3. 3

    Ask. Inspect. Understand.

    Ask in plain language. Read the commands and their output, then follow up in the same conversation.

The answer matters. So does how you got there.

An investigation shouldn't disappear into a pile of terminal tabs. Keep the question, the command and the result together.

  • Inspect the actual output

    Read command results alongside the explanation, including failed attempts and retries.

  • Keep the thread

    Return to a previous conversation and pick up the investigation with its history intact.

  • Give each role its place

    Members use read-only cluster tools. Organization admins and owners can manage resources.

An investigation you can follow Illustrative output
kubectl describe pod api-7c9f -n production 
Containers:
  api:
    Last State:     Terminated
      Reason:       OOMKilled
      Exit Code:    137
    Restart Count:  12
    Limits:
      memory:       128Mi

The useful clue is right here: the container ran out of memory. Check usage before adjusting the limit.

Saved with the conversation

A plan for your cluster. And your team.

Start with what you run today. Add room as your infrastructure grows.

Individual

For personal use.

€20 / month

14-day free trial

Start free trial
  • 2 Kubernetes clusters
  • 2 team members
  • AI usage budget included
  • Chat and command history
  • Roles for your organization

Enterprise

For large teams and enterprises.

€499 / month

For larger organizations

Get started
  • 20 Kubernetes clusters
  • 50 team members
  • AI usage budget included
  • Chat and command history
  • Roles for your organization

Need more AI usage? Add credit to your organization's budget.

A clearer next step starts here.

Practical Kubernetes guides for the moments when the error message isn't enough.

More from the blog

Before you connect.

The practical questions, answered.

Ask us something else
How does my cluster connect?

Register the cluster in your organization and install its generated agent manifest. The agent opens the connection to Ranching.Farm and returns command results to your chat.

Can the assistant change my cluster?

Members have read-only cluster tools. Organization admins and owners can run modifying commands. Check each proposed command carefully: changes affect your live workloads.

Which tools does it use?

The cluster assistant works with kubectl, Helm and Kustomize. It can explain their output and turn cluster relationships into diagrams.

Can I return to an earlier investigation?

Yes. Your chat history keeps the conversation, tool results and command attempts together, so you can continue with the previous context.

Can I try it before connecting my own cluster?

Yes. The public Kubernetes demo is a separate environment. Individual and Team plans also include a 14-day trial for working with your own clusters.

Your next incident deserves a second pair of eyes.

Try a question in the demo. Connect your own cluster when you're ready.

Try the Kubernetes demo