Flash⚡️Learn

✨ About the Project: Flash⚡️Learn

Tagline: From learning to revising — all in a flash.

🚀 Inspiration

The idea for Flash⚡️Learn came from a real-world problem — revisiting your own work before an interview or exam. Imagine having to quickly brush up on a project you completed months ago. Reading lengthy reports or notes isn’t practical when time is short. What truly helps is answering flashcard-style questions, which let you absorb the core concepts and refresh your memory fast.

We also saw an opportunity to help learn new topics on-the-go. Whether it's brushing up on machine learning, DevOps, or even world history — why not get the key takeaways instantly in a flashcard format?

Flash⚡️Learn was born from this dual need:
Revisit old knowledge or learn something new — instantly.

💡 What it does

Flash⚡️Learn is an AI-powered flashcard generator that offers two core modes:

Revise from your notes or documents — Upload a PDF or paste your notes, and Flash⚡️Learn will instantly generate intelligent flashcards to help you recall key points.
Learn a brand-new topic — Just enter a topic like “Blockchain” or “Kubernetes” and get AI-generated Q&A flashcards to quickly get up to speed.

All flashcards are beautifully presented with a flip-card UI for smooth, interactive studying.

🛠️ How we built it

Frontend:
Built with React + Vite, styled using Tailwind CSS, and enhanced with Font Awesome icons for a modern and accessible UI. Interactive flip-cards provide an intuitive flashcard experience.
Backend:
Developed using FastAPI (Python).
- PDF files are processed using PyPDF2 to extract raw text.
- The text or user-given topic is sent to OpenAI’s GPT API, which returns flashcards in structured JSON format.
APIs & Integration:
- OpenAI GPT-3.5 Turbo (for generating flashcards)
- Render is used to host and deploy the backend server.

🧩 Challenges we ran into

Crafting the right prompt:
Getting consistently structured and relevant output from the GPT API required careful prompt engineering.
PDF text extraction quirks:
Not all documents are well-formatted, which made it challenging to extract clean, usable text.
GPT formatting inconsistencies: Getting consistent JSON output from the GPT API was tricky. We had to parse and validate responses properly on the backend to ensure our frontend didn’t crash.

🏆 Accomplishments that we're proud of

Solving a real-life problem:
We tackled a situation that many students and professionals face — revisiting knowledge efficiently — and delivered a practical solution.
Full-stack integration:
Successfully connected frontend, backend, and AI into a smooth, end-to-end experience within a limited timeframe.
Dual-mode functionality:
Supporting both learning and revision workflows while keeping the UI intuitive and clean.

📚 What we learned

The importance of prompt design in LLM-based applications.
Full-stack web app building with React, Tailwind, FastAPI, and external APIs.
Deploying backend services using Render, and managing cross-origin frontend/backend communication.