Codegraph - Brains for your AI - Graph-based code intelligence that understands your codebase

CodeGraph is an open-source code intelligence system that transforms your code into a queryable graph database. Instead of parsing files on-the-fly, it indexes your codebase into structured relationships—classes, functions, dependencies, all connected and searchable in milliseconds. Perfect for powering AI assistants, building smarter code editors, or analyzing large codebases at scale. Start with Python. Expand to any language. Own your data. No vendor lock-in.

We started this project because we kept hitting the same wall: AI models and code analysis tools were flying blind. They'd parse files on-demand, lose context across the codebase, or get overwhelmed by large projects. Meanwhile, databases like Neo4j existed, but nothing was purpose-built for code. So we built a code-aware graph database from scratch. The core idea is simple: index your codebase once into a persistent graph, then query it in <100ms. This powers smarter AI context, enables real code analysis at scale, and gives developers instant answers about their own code. What we're excited about: The VS Code extension launching soon (LSP-based, with AI context provision) Expanding beyond Python to Rust, JavaScript, and beyond The potential for AI assistants to actually understand your specific codebase instead of hallucinating We'd love your thoughts on: What languages should we prioritize next? How would you use a code intelligence system in your workflow? Any thoughts on the open-source + SaaS model approach? All code is open on GitHub under Apache 2.0—no gatekeeping on the foundation. We're building this in the open because good code intelligence benefits everyone. Happy to answer any questions! 🚀

Codegraph - Brains for your AI - Graph-based code intelligence that understands your codebase

Replies