Quiz & Viva Arena

An AI-powered educational application for generating, reviewing, and evaluating flashcards. Designed to enhance studying with LLM-assisted answer validation and targeted feedback.

Features

• AI Flashcard Generation: Ingests raw notes and leverages Claude 3 to generate concise, exam-style flashcards.
• Premade University-level Decks: Instantly generate comprehensive flashcards for topics like Human Anatomy II, Microbiology, Organic Chemistry, and Calculus I.
• Intelligent Answer Checking: Free-text answers are evaluated by AI, providing an accuracy score and brief, encouraging feedback based on the core meaning, rather than exact string matching.
• AI Quiz Generation: Upload your study materials to automatically generate interactive quizzes with both multiple-choice and open-ended questions using Gemini.
• Interactive Quizzes & AI Grading: Complete quizzes interactively with live AI evaluation, scoring, and detailed feedback for open-ended answers.
• Dynamic Frontend: Modern UI with rating mechanisms to assess confidence levels for active recall and spaced repetition.

Tech Stack

• Frontend: React 19, React Router, TailwindCSS 4, Vite
• Backend: FastAPI, SQLAlchemy (SQLite), Pydantic
• AI Models: Anthropic Claude 3 Haiku & Google Gemini 3.0 Flash

Prerequisites

• Node.js (v18+)
• Python (3.9+)
• Anthropic API Key (Claude)
• Google Gemini API Key

Getting Started

1. Setting up the Backend

1. Open a terminal and navigate to the backend directory:

   cd src

2. Create and activate a virtual environment (optional but recommended):

   python -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install the required dependencies:

   pip install -r requirements.txt

4. Create a .env file in the project root with your API key:

   GEMINI_API_KEY="your_google_gemini_api_key_here"
   GEMINI_MODEL="gemini-3-flash-preview"
   CLAUDE_API_KEY="your_anthropic_api_key_here"
   CLAUDE_MODEL="claude-3-haiku-20240307"

5. Start the FastAPI server:

   uvicorn main:app --reload

   The backend will start and be available at http://localhost:8000. You can visit http://localhost:8000/docs to test the API endpoints using Swagger UI.

2. Setting up the Frontend

1. Open a new terminal and navigate to the client directory:

   cd client

2. Install Node dependencies:

   npm install

3. Start the Vite development server:

   npm run dev

   The frontend will be available at http://localhost:5173. Make sure the Python backend is running so the frontend can properly call APIs.

Folder Structure

• /src - FastAPI application, routers, database models, and Claude AI integration logic.
• /client - React frontend powered by Vite, utilizing Tailwind for responsive styling.