Cairn Documentation
What is Cairn?
Cairn is an AI-native platform for Model-Based Systems Engineering. It helps engineers turn rough system concepts into structured, traceable models — the kind of structured thinking that used to require expensive enterprise tools or weeks of manual documentation.
The name comes from the stone markers hikers build to mark a path through complex terrain. Each stone is placed deliberately, one at a time, creating something that guides others through uncertainty. That's what Cairn does for system design.
What Cairn is For
Cairn is built for the "thinking phase" of engineering — the messy early work where you're figuring out what a system actually needs to be. You describe a system in plain language, and Cairn helps you decompose it into subsystems, define requirements, model behavior, and track design decisions. The AI proposes structure; you review and refine.
This is the phase that typically happens in spreadsheets, slide decks, and whiteboard photos — artifacts that don't connect to each other and can't answer basic questions like "what depends on this?" or "why did we reject that approach?"
What Cairn is Not
Cairn is not a replacement for enterprise MBSE tools like Cameo, DOORS, or Rhapsody. Those tools are designed for formal verification, regulatory compliance, and large-team coordination on multi-year programs. Cairn is designed for the work that happens before you need those tools — or for the many projects that will never need them at all.
If your system will eventually need SysML v2 diagrams or ReqIF exports for regulatory submission, Cairn gives you a structured starting point to hand off. If your system is a prototype, a side project, or an early-stage product, Cairn might be all you need.
Your Data, Your Machine
Your model lives in your browser's local storage and saves as you work. It reaches Cairn's servers only for a project you back up to your account, which is a switch for each project in Settings. A project you start in the interview backs up as you go, so you can carry on from another device: on a computer you can untick it, and an interview begun on a phone always backs up, since a computer is where you finish it.
AI features need a network connection: the part of the model a request needs is sent to Claude. You can export a whole project at any time.