An AI-powered arXiv search engine. Find, understand, and save scientific papers in seconds.
- 🔍 Manual search Type a keyword and explore relevant arXiv papers.
- 🤖 AI-powered search Ask questions in natural language and get smart results.
- 📊 Vector Search with MongoDB Results improve with use.
- 🌐 Blazing fast frontend Built with Next.js 15 + Vercel.
- 📚 arXiv integration Live access to the latest research.
- 🧠 AI-powered paper summarizer (TL;DR in one click)
- 🧾 Save articles with user accounts
- 🪄 A sleeker UI
- ⚡️ Backend performance boost
- 🌍 Custom domain
| Layer | Tech |
|---|---|
| Frontend | Next.js 15 (App Router) |
| Backend | FastAPI + Next.js API routes |
| Database | MongoDB Atlas + Vector Search |
| AI Search | sentence-transformers (MiniLM-L6-v2) |
| Embedding | Hugging Face 🤗 + MongoDB Atlas Vector |
| Infra | Vercel (Frontend) + Railway (Backend) |
| AI SDK | Groq x Vercel SDK |
Alan Geirnaert
MIT License. Use it, fork it, launch your own scientific tools.
If you like the project, star ⭐ it on GitHub, share it, or contribute!