Diagrams-as-Code with AI
More and more teams are starting to use coding to tackle their diagramming and documentation challenges. Using plain text notations to create diagrams turns a traditionally artistic process into a systematic and data-driven approach. Using diagrams-as-code and models instead of traditional diagrams gives you many benefits, including storing your diagrams alongside your code, automatic consistency, and the ability to engage AI in the process.
Join this masterclass to explore the more popular diagrams-as-code notations so that you can create diagrams-as-code as well as understand the diagrams-as-code that an LLM might generate for you. We all know that the output of an LLM cannot be trusted, so you need to understand the code the LLM supplies to you.
We will cover PlantUML, Mermaid, and Structurizr, for use alongside your code and in documentation and build pipelines. We will also delve into the use cases and limitations of using these diagrams-as-code with AI. All the tools we use in this workshop will be free to use.
Agenda
Introduction
Diagrams-as-code notations:
- Comparing PlantUML, Mermaid, & Structurizr
- Diagrams vs. diagrams-as-code vs. models
PlantUML:
- Diagrams, syntax, and tools
- Using PlantUML with LLMs
- Exercises
Mermaid:
- Diagrams, syntax, and tools
- Using Mermaid with LLMs
- Exercises
- Comparing PlantUML & Mermaid
Structurizr:
- Diagrams, syntax, and tools
- Using Structurizr with LLMs
- Exercises
- Comparing PlantUML, Mermaid, & Structurizr
Wrap Up
Audience
Developers, Architects, Tech Leads, Technical Writers, Business Analysts, anyone wanting to adopt diagrams-as-code.
Learning Outcomes
- Gain a grounding in PlantUML, Mermaid, and Structurizr
- Understand the tradeoffs of diagrams-as-code and models vs. standard diagrams
- Recognise the limitations and useful use cases of using diagrams-as-code with LLMs
- Learn through hands-on exercises and interactive activities
