How govindmehta/studyzen works
Project type and purpose
• This is a full-stack “StudyZen” web application acting as an AI-powered study companion.
• Frontend: React/TypeScript SPA built with Vite and Shadcn UI.
• Backend: FastAPI Python service leveraging LLMs (Google Gemini via LangChain), ChromaDB vector store, video & PDF processing, and auxiliary scripts.
File structure
• Top-level directories:
• /frontend – all UI code, components, pages, static assets, config (Vite, Tailwind, ESLint) and Prisma ORM schema/migrations.
• /backend – FastAPI entrypoint (main.py), routes (quiz.py, quiz.py, flowchart.py, etc.), AI integrations (rag.ipynb, graphvizzz.py, content.py), DB files (ChromaDB folder and SQLite), PDF & TTS utilities, requirements.txt.
• Patterns:
• Separation of concerns between UI and API.
• ORM migrations under frontend/prisma represent user auth & flashcard schemas.
• AI-centric modules for explanation, flowchart generation, quizzes, and TTS.
• Config & deployment:
• frontend/.env for VITE_API_URL
• backend/.env for GEMINI_API_KEY, YOUTUBE_API_KEY
• uvicorn startup in README.
README insights
• Tech stack calls out React, FastAPI, Prisma, ChromaDB, LangChain, OpenCV, Pandas.
• API endpoints: /explanation/, /generate_schedule/, /flowchart/, /videos/
• Database: Prisma (Postgres/SQLite) for user & progress, ChromaDB for vector embeddings.