PrepAxis

An AI-powered placement preparation platform for college students. Practice mock interviews with real-time AI feedback, get questions generated around your own resume, and browse a searchable interview question bank.

🔗 Live demo: prepaxis-pi.vercel.app

The backend runs on Render's free tier and spins down after periods of inactivity — the first request after idle time can take up to ~50 seconds to wake up. Subsequent requests are fast.

Try it instantly — no signup required

Email	Password
test1@example.com	password123

---

Features

• 🎯 Resume-aware mock interviews — upload a PDF/DOCX resume and get interview questions generated around your actual experience, or practice without one
• 💬 Real-time AI feedback — each answer is scored and critiqued as you go, powered by Google Gemini
• 📊 Dashboard — track past interview attempts, scores, and recurring weak topics over time
• 📚 Question bank — browse interview questions by topic and difficulty; generate new AI questions and save your own personal additions alongside the shared bank
• 🔐 Authentication — JWT-based signup/login with hashed passwords

Architecture

A React/TypeScript frontend sends JWT-authenticated API calls to a Node.js/Express backend. The backend is organized into five modules — Auth, Interview, Resume, Question bank, and Dashboard — which read and write to MongoDB and call Google Gemini for question generation and feedback.

Tech Stack

Frontend	React 19, TypeScript, Vite, Tailwind CSS 4, React Router 7
Backend	Node.js, Express 5, MongoDB (Mongoose)
AI	Google Gemini (Flash)
Auth	JWT, bcrypt
File parsing	pdf-parse, mammoth (résumé text extraction)
Deployment	Vercel (frontend) · Render (backend) · MongoDB Atlas (database)

Project Structure

backend/

• controllers/ — Request handlers (auth, interview, question, resume, dashboard)
• models/ — Mongoose schemas (User, InterviewSession, Question, Resume, SavedQuestion)
• routes/ — Express route definitions
• middleware/ — Auth guard, file upload, AI rate limiting
• services/ — Gemini API integration, resume text extraction
• prompts/ — AI prompt templates
• scripts/ — Demo user seed script
• server.js

frontend/src/

• pages/ — Login, Signup, Dashboard, Interview, QuestionBank, Resume
• context/ — Auth context
• api/ — Axios client

Running Locally

Prerequisites

• Node.js (v18+)
• A MongoDB connection string free tier works fine, or a local instance
• A Google Gemini Key

1. Clone the repo

git clone https://github.com/anushka12060/Prepaxis.git
cd Prepaxis

2. Backend setup

cd backend
npm install
cp .env.example .env

Fill in .env with your real values: MONGO_URI=your-mongodb-connection-string JWT_SECRET=any-long-random-string GEMINI_API_KEY=your-gemini-api-key PORT=5000

Start the server:

npm run dev

3. Frontend setup

In a new terminal:

cd frontend
npm install
cp .env.example .env

Fill in .env: VITE_API_URL=http://localhost:5000/api

Start the dev server:

npm run dev

Open http://localhost:5173.

4. (Optional) Seed a demo account

cd backend
npm run seed:demo

Creates/resets a demo user you can log in with — test1@example.com / password123.

API Overview

Route	Description
POST /api/auth/signup · /api/auth/login	Authentication
POST /api/resume/upload	Upload and parse a resume (PDF/DOCX)
GET /api/resume	Retrieve latest resume
POST /api/interview/start	Start a new mock interview session
POST /api/interview/:id/answer	Submit an answer, get AI feedback + next question
POST /api/interview/:id/complete	Finish the interview, get an overall score and summary
GET /api/questions	Browse the question bank
POST /api/questions/generate	Generate a new AI question
GET /api/dashboard	User's interview stats and history

All routes except signup/login require a Bearer JWT token.

License

MIT