Inspiration

Understanding an unfamiliar codebase can be frustrating and time-consuming. We wanted to create a way for developers to explore complex codebases through natural language instead of manually tracing files and dependencies.

What We Built

AXT Graph transforms a codebase into a structured knowledge graph connecting files, functions, classes, and their relationships. Developers can then ask questions about the codebase and get contextual answers based on its structure.

How We Built It

We built AXT Graph by analyzing code and extracting meaningful entities and relationships, organizing them into a queryable graph, and connecting that graph to an AI-powered natural-language interface.

What We Learned

We learned how important structured context is when using AI for software engineering. Giving an AI the relationships within a codebase can make its answers much more useful than simply providing raw files.

Challenges

Our biggest challenge was turning messy, interconnected source code into a useful structure while keeping the graph understandable and the AI responses relevant.

AXT Graph's goal is simple: make complex codebases easier for both developers and AI to understand.