Course

Vibe Modelling Short

A practical short guide to get you started with vibe modelling!

Minutes

Video

What will you learn?

 
Technical diagrams are essential for analysts and architects, but they can be time-consuming to create and maintain. In this short video, you will learn how vibe modelling lets you use an LLM in conversation to produce diagram descriptions as code, which a renderer such as PlantUML, Mermaid, or Graphviz can reliably turn into clean visuals. You will also see why asking an LLM to generate the visual directly often leads to incorrect diagrams, and how combining an LLM with a renderer improves quality and makes iteration faster.

Focus on the “why”

Many AI generated diagrams look convincing at first glance, but contain logical and notation errors. You will learn the core reason this happens and when you can and cannot rely on direct visual output.

Use code for diagramming

Instead of generating a picture, you generate diagram code. You will learn how text based diagramming languages capture structure and semantics, while the renderer enforces rules and layout.

Two ways to iterate

You can refine a diagram by updating your prompt and regenerating the code, or by editing the code directly in the rendering tool. You will learn when to use each approach for speed and precision.
Tools that let you chat to diagrams

Tools that let you chat to diagrams

You will get a practical overview of common workflows and tools, including renderers like PlantUML and Graphviz, and AI enabled diagram tools that integrate chat directly, such as Eraser.io.