secondary case study
Codebase Genius
AI documentation generator
- Status
- Live demo
- Engagement
- Independent project
- Role
- Full-stack & AI engineer
The problem
GitHub repositories often lack comprehensive, up-to-date documentation, making onboarding and maintenance slow.
What was built
A tool that generates documentation from a GitHub repository using LLM analysis.
Audience
Development teams and open-source maintainers.
Technical highlights
- Repository analysis with Google Gemini
- Automated documentation generation
- Graph-based code relationships (NetworkX)
My responsibilities
- Product design
- Full-stack implementation
Why it matters for a client
Shows code analysis, graph modelling and LLM-assisted documentation.
Constraints and trade-offs
Demo application; output quality depends on repository structure and the LLM provider.
Privacy and accessibility
Reads public repository code; does not store private source unless explicitly provided.
Verified evidence
Live demo at geniuscodebase.streamlit.app; public source at github.com/rayklanderman/jaseci-proj.
Limitations
Demo scope; not a production documentation pipeline.
Technology
PythonGoogle GeminiStreamlitNetworkX