Inspiration
Born out of countless all‑nighters and the chaos of last‑minute studying, this project tackles a simple pain: when time is short, decision‑making is the enemy. “What format should I study in? Where do I even start?” So, this tool provides a single, focused entry point that instantly spins your material into three productivity modes — a cheatsheet, flashcards, or a Keynote‑style presentation — letting you dive straight into learning instead of planning.
The goal: help you prepare an exam in two hours or less with tools that balance speed, clarity, and minimal cognitive friction. You can still edit details where needed without breaking the rest of your workflow — no domino effect of formatting disasters.
What it does
Home Page: A clean, centered textarea welcomes your input, with the ability to upload .jpg, .jpeg, .txt, or .pdf files. From there, pick your mode — cheatsheet, flashcards, or presentation — and jump in.
Cheatsheet Page: A full LaTeX editor with real‑time compile and PDF preview, plus an integrated chat panel for quick, interactive fine‑tuning. Write, render, iterate — all without leaving the page.
Flashcard Page: A JSON editor and deck loader with an intuitive flip‑card study mode. Type a concept, flip for the answer, and repeat until panic transforms into calm mastery.
Challenges we ran into
- PDF ↔ LaTeX Mappings: Sparse documentation meant months condensed into hours of exploratory trial and error to build a stable, reversible link between LaTeX and rendered PDF.
- Performance juggling: Keeping the LaTeX editor, PDF renderer, and chat in sync without freezing the browser required careful worker threading and UI isolation.
- Rate Limits T^T: Novel solution doesn't mean it is feasible for production huhuhu, at the end of the day deployment > theory.
What's next for I am beyond cooked and my exam is in 2 hours!
Strengthen LaTeX–PDF round‑trip reliability. Add smart AI summarization that still lets users control edits without unexpected rewrites. Refine flashcard analytics and spaced repetition for better long‑term memory retention. Maybe… sleep.
Built With
- atlas-rag
- fastapi
- latex
- openai
- python
- react
- typescript
- vite
Log in or sign up for Devpost to join the conversation.