Knowledge in code is gold — how to extract it and power your team and AI
summary
Talk demonstrates techniques to extract knowledge from code to feed teams and AI, comparing tools and approaches: GitDiagram (which dumps full repos into prompts but gives unstable diagrams), Cursor (which finds flows but hallucinates handler class names), and DeepWiki paired with the Devin coding agent to query a generated wiki. It also covers convention-based extraction using Structurizr/C4 and Spring Modulith for repeatable component models, plus Noesis.vision’s P3 domain model for stable documentation and Upmap’s use of recorded integration tests to produce runtime sequence traces.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 6:04Automating code docs with AI
- 17:24AI role in docs and limits
- 22:05Static code analysis tools
- 29:54P3 model and docs
- 36:02Runtime code analysis uses
- 39:42Summary
- 42:32Q&A
related talks
38mCode reading
44mAutomate your software design How far we can get and how the AI chan
36mCode as Data: How data insights on legacy codebases fill the knowledge gap in modernization projects
28mI've Fallen Into a Large Codebase and I Can't Get Up!
47mExploring a complex codebase with AI
40mContext That Dreams
55mDigging into the Matrix: Practicing Code Archaeology
38mAI for Software Development: A Reality Check