PrepFlow 🚀

A Production-Ready AI-Powered Interview Preparation Platform

PrepFlow is a robust, full-stack web application designed for software engineers to master data structures, algorithms, and system design concepts. It combines a modern Kanban workflow, a mathematically rigorous Spaced Repetition System (SM-2), and an AI-powered Mock Interview evaluator.

---

✨ System Architecture

                  +-------------------+
                  |   React Frontend  |
                  | (Vite, Redux, Dnd)|
                  +--------+----------+
                           |
                           | REST API (JWT Auth)
                           v
                  +--------+----------+
                  |  FastAPI Backend  | 
                  |  (Rate Limiting,  | 
                  |   Pydantic Val.)  |
                  +--------+----------+
                           |
       +-------------------+-------------------+
       |                   |                   |
       v                   v                   v
+-------------+     +-------------+     +-------------+
|   MongoDB   |     | SM-2 Engine |     | AI Engine   |
| (Motor Async|     | (Spaced Rep)|     | (Mock Intv.)|
+-------------+     +-------------+     +-------------+

📸 Screenshots

 

 

---

🔥 Features & Capabilities

Feature	Description
🤖 AI Mock Interviews	Simulated, timed interviews. Enter your solution and receive AI-driven scoring (0-100), strength analysis, and improvement feedback.
🧠 True SM-2 Spaced Repetition	Uses the formal SuperMemo-2 algorithm. Interval and Ease Factor (EF = EF + (0.1 - (5-q)*(0.08+(5-q)*0.02))) computed on the backend.
📊 Analytics Insight Engine	Goes beyond charts. Computes your weakest topics, hardest difficulty drop-offs, and 30-day consistency score.
📋 Kanban Board	Highly optimized drag-and-drop board with column grouping, due-date pulse animations, and optimistic UI updates.
🛡️ Backend Hardening	Global in-memory rate limiting, Request logging middleware, Pydantic input validation, and MongoDB indexing.

---

📸 Core Pages

Page	Route	Description
Board	/	Kanban drag-and-drop board
Topics	/topics	All topics with progress tracking
Topic Detail	/topics/:id	Debounced search and difficulty filtering
Reviews	/reviews	SM-2 Spaced repetition review session
Analytics	/analytics	Real-time charts, GitHub-style heatmaps, and AI insights
Mock Interview	/mock-interview	Timed interview simulator with AI grading

---

🚀 Deployment Guide

This project is built for modern serverless and PaaS deployment.

1. Frontend (Vercel)

1. Push your repository to GitHub.
2. Go to Vercel and import the project.
3. Set the Framework Preset to Vite.
4. Set the Root Directory to frontend.
5. Add Environment Variable:
   • VITE_API_URL=https://your-backend-url.onrender.com

2. Backend (Render / Railway)

1. Create a new Web Service on Render/Railway.
2. Connect the repository and set Root Directory to backend.
3. Build Command: pip install -r requirements.txt
4. Start Command: uvicorn main:app --host 0.0.0.0 --port $PORT
5. Environment Variables:
   • MONGODB_URL: Your MongoDB Atlas connection string.
   • SECRET_KEY: A strong secure random string.
   • CORS_ORIGINS: https://your-vercel-app.vercel.app

3. Database (MongoDB Atlas)

1. Create a free cluster on MongoDB Atlas.
2. Go to Database Access and create a user (username/password).
3. Go to Network Access and allow IP 0.0.0.0/0.
4. Copy the connection string into your Backend Environment Variables.

---

💻 Local Setup Instructions

1. Clone the repository

git clone https://github.com/1tsadityaraj/PrepFlow.git
cd PrepFlow

2. Backend Setup

cd backend
python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate
pip install -r requirements.txt
# Run the server (auto-seeds mock data in-memory for testing)
uvicorn main:app --reload

3. Frontend Setup

cd frontend
npm install
npm run dev

The app will be running at http://localhost:5173.

---

📄 Resume Bullet Points

• Architected and developed a full-stack interview preparation platform using React, Tailwind CSS, FastAPI, and MongoDB, achieving a robust, scalable architecture with global rate limiting and comprehensive Pydantic request validation.
• Implemented a mathematically rigorous Spaced Repetition System (SM-2) on the backend, tracking User Ease Factor and optimal intervals to maximize long-term retention of algorithmic concepts.
• Integrated an AI-powered Mock Interview Simulator, enabling users to take timed, randomized assessments and receive instantaneous, multi-faceted feedback (score, strengths, weaknesses) via a custom evaluation engine.
• Engineered an Analytics Intelligence Layer that processes 30-day user activity data to render GitHub-style heatmaps, identify weak topics, calculate consistency scores, and provide actionable study insights using Recharts.
• Optimized frontend performance by implementing debounced search inputs, useMemo/React.memo for component virtualization, and robust Redux state management across a highly interactive drag-and-drop Kanban workflow.