Platform Documentation
Platform Documentation
Breadcrumbs

RunWhen Platform Documentation

RunWhen is an AI SRE platform that automates troubleshooting and remediation for Kubernetes and cloud environments. It works in two modes:

  • Autonomous mode — RunWhen continuously runs diagnostic tasks in the background, building production insights: structured findings about the health, configuration, and behavior of your services. These insights are always current and ready for the AI to reference — no waiting for data collection when something goes wrong.

  • Interactive mode — When you ask a question in Workspace Chat (or an alert fires), an AI assistant combines those background insights with new, targeted diagnostics to give you an actionable answer — root cause, remediation steps, or what to investigate next.

All tasks execute inside your own clusters via a lightweight agent. Your credentials and data stay in your network; the platform coordinates what to run.

Get Started

🏁 Try It Now

Jump into the Live Demos — no setup required. See how the platform diagnoses real application failures in a shared sandbox.

🚀 Set Up Your Environment

Ready to connect your own clusters? The Install section walks you through deployment in about 30 minutes.

📚 Learn How It Works

Want to understand the architecture first? Learn covers the key concepts — workspaces, tasks, assistants, and more.

Day-to-Day Usage

Once your environment is connected, the Use section covers common workflows: asking questions in Workspace Chat, reviewing issues, running tasks, setting up rules and commands to customize your assistants, and sharing findings with your team.

Documentation Map

Live Demos

Hands-on tutorials using a shared sandbox workspace. Walk through real failure scenarios and see AI-assisted diagnosis in action.

Learn

Core concepts: workspaces, assistants, tasks, issues, and how the system discovers and monitors your infrastructure.

Install

Deployment guides for SaaS with a private runner, or self-hosted. Includes prerequisites, helm chart configuration, and verification steps.

Configure

Post-install setup: cloud discovery, SSO, user management, Slack integration, webhooks, secrets, and custom task configuration.

Use

Workspace Chat, Workspace Studio (tasks, rules, commands, knowledge), issue management, and everyday platform workflows.

Architecture & Overview

System design, data flow, deployment topology, and scalability considerations.

How It Works (30-Second Version)

  1. You ask a question (“Why is checkout failing in production?”) or an alert fires.

  2. An AI assistant picks relevant tasks from a library of expert-authored automation.

  3. Tasks run in your environment via a lightweight agent (the “runner”) that stays inside your cluster.

  4. Results are analyzed and returned with root-cause findings and recommended next steps.

Your credentials and cluster data never leave your network. The platform coordinates what to run; the runner executes it locally.

Platform & Tools

Support

Additional resources: For task authoring and CodeBundle development, see the RunWhen Authors Documentation