I built a small AI-assisted tool that predicts which graph algorithm (e.g., Dijkstra, BFS, coloring) to apply based on graph properties.

GitHub: https://github.com/lixiasky/Graphtune Model: XGBoost trained on synthetic + real graph data Use Case: Automatically selects the fastest algorithm for your input graph.

---

But the reason I built it might be more interesting.

A few days ago, I watched this video by a brilliant girl from Imperial College math department. She explained a Chinese math olympiad problem using Manim animations and really deep insights: https://www.bilibili.com/video/BV1WsMCzeEQg/

Unfortunately, the video didn’t get much attention, and she mentioned on stream that she cried after seeing the low views. So I made this project inspired by her problem and her visuals, hoping to help more people discover it.

If you like this idea, you don’t need to star my repo — but please consider clicking into her video, giving it a thumbs up, and leaving a kind comment. Maybe just a simple “nice work” will make a big difference.

Thanks