Builder working on LLM applications, memory systems, graph-based RAG, and multimodal AI.
MS ECE, Georgia Tech. Previously did algorithm engineering work at Tencent and Baidu.
I’m building Understand Anything, an open-source tool that turns any codebase into an interactive knowledge graph for exploration, search, and plain-English understanding.
I am working on Understand-Anything, trying to use AI to teach you how to understand something. The problem for me, as a beginner in lots of aspect, is do not know what question to ask to AI in a area that I am not familiar with.
One thing I realized while working on large repos is that most “code graph” tools are still fundamentally navigation tools.
You can see structure, dependencies, call graphs, etc., but you still spend a lot of time manually building a mental model of why things exist and how concepts connect across the codebase.
What I’m trying to explore with Understand Anything is whether LLMs + structured graphs can help generate higher-level semantic understanding instead of only visualization.
For example:
1. tracing how a business concept propagates through services/modules
2. mapping requirements ↔ implementation ↔ data flow
3. surfacing architectural patterns automatically
4. helping new contributors build a mental model faster
Still very early obviously, but that’s the direction I’m interested in exploring.
I use it in my company, a startup that vibe coding for thousands line of code every day. And it works fine. Engineers love the business knowledge mode.
I built Understand Anything to help with a problem I run into often: understanding an unfamiliar codebase without spending hours jumping between files, docs, and dependency paths.
The tool analyzes a repo and builds an interactive map of files, functions, classes, dependencies, and architectural relationships. You can explore the map, search across the project, ask questions, generate onboarding notes, and inspect what may be affected by a change.
It started as a Claude Code plugin, but I’ve been making it work across different AI coding environments such as Codex, Cursor, Copilot, Gemini CLI, OpenCode, and others.
The main design goal is to avoid producing a giant graph that looks impressive but does not actually help people understand the code. I am trying to make the graph useful as a learning and navigation layer.
I’d appreciate feedback from people who work on large repos, onboard new engineers, or use coding agents.
Graphs that teach > graphs that impress. Turn any code, or knowledge base (Karpathy LLM wiki), into an interactive knowledge graph you can explore, search, and ask questions about. Works with Claude Code, Codex, Cursor, Copilot, Gemini CLI, and more.
MS ECE, Georgia Tech. Previously did algorithm engineering work at Tencent and Baidu.
I’m building Understand Anything, an open-source tool that turns any codebase into an interactive knowledge graph for exploration, search, and plain-English understanding.