GapWise AI 🎯

RAG & AI-powered interview preparation platform — upload your resume, paste the job description, and get a personalized interview report, predicted questions, skill gap analysis, and a tailored resume PDF in seconds.

---

✨ Features

• 🔐 JWT Authentication System
• 📄 Resume Upload and PDF Parsing
• 🧠 RAG-Powered Resume Matching Pipeline
• 🤖 AI Interview Report Generation (Gemini + Zod schema)
• 🤖 AI Interview Report Generation (Zod schema)
• 📊 Match Score Normalization and Skill Gap Analysis
• 💬 Technical and Behavioral Questions with intent and answer guidance
• 📅 Personalized Preparation Roadmap
• 📥 ATS-Friendly Resume PDF Generator
• 🪟 Windows-safe PDF generation fallback (local Chrome/Edge support)
• 📚 Interview Report History
• 🐳 Dockerized Full Stack Application
• ⚙️ GitHub Actions CI/CD Pipeline
• 🧪 Automated API Testing with Jest and Supertest

---

🛠️ Tech Stack

Backend

Technology	Purpose
Node.js + Express	REST API server
MongoDB + Mongoose	Database
JWT + bcryptjs	Authentication
Google Gemini AI via @google/genai	Interview report and resume HTML generation
Local RAG service	Resume chunking, embedding, retrieval
Puppeteer Core + Chromium	HTML to PDF conversion
Multer	Resume upload handling
pdf-parse	PDF text extraction
Zod + zod-to-json-schema	AI output schema enforcement
Jest + Supertest	API tests

Frontend

Technology	Purpose
React 19 + Vite	Frontend framework
React Router v7	Routing
Axios	API requests
SASS	Styling

DevOps and Deployment

Technology	Purpose
Docker	Containerization
Docker Compose	Multi-container orchestration
GitHub Actions	CI/CD automation
Nginx	Frontend production server
Render	Deployment platform

---

🚀 Local Development Setup

Prerequisites

• Node.js >= 18
• Docker and Docker Compose
• MongoDB Atlas account or local MongoDB
• Google Gemini API key
• Pinecone account and API key

---

1. Clone Repository

git clone https://github.com/Akshay4754/GapWiseAI.git
cd GapWiseAI

---

2. Backend Setup

cd Backend
npm install

Create .env inside Backend/

PORT=3000
MONGO_URI=your_mongodb_connection_string
JWT_SECRET=your_jwt_secret
GOOGLE_GENAI_API_KEY=your_gemini_api_key
PINECONE_API_KEY=your_pinecone_api_key
PINECONE_INDEX_NAME=gapwise-ai-rag-google
GOOGLE_EMBEDDING_MODEL=gemini-embedding-001
GOOGLE_EMBEDDING_DIMENSIONS=768
FRONTEND_ORIGIN=http://localhost:5173,http://localhost:5174

Run backend

npm run dev

If npm run dev has trouble with nodemon, run:

npm start

---

3. Frontend Setup

cd Frontend
npm install

Create .env inside Frontend/

VITE_API_URL=http://localhost:3000

Run frontend

npm run dev

Frontend runs at

http://localhost:5173

---

🐳 Docker Setup

Run Full Application

docker-compose up --build

Containers expose

• Frontend on http://localhost:3000
• Backend on http://localhost:5000
• MongoDB on localhost:27017

Build Containers Individually

Backend

docker build -t gapwise-backend ./Backend

Frontend

docker build -t gapwise-frontend ./Frontend

---

⚙️ CI/CD Pipeline

GitHub Actions workflow in .github/workflows/ci.yml automatically

• ✅ Installs backend and frontend dependencies
• ✅ Runs Jest + Supertest tests
• ✅ Builds Docker images
• ✅ Pushes Docker images on push to main, only if tests pass

---

🔐 Required GitHub Secrets

Go to GitHub Repository → Settings → Secrets and variables → Actions

Add

DOCKER_USERNAME
DOCKER_PASSWORD
MONGO_URI_TEST
GEMINI_API_KEY

---

📁 Updated Project Structure

GapWiseAI/
├── .github/
│   └── workflows/
│       └── ci.yml                      # GitHub Actions CI/CD pipeline
│
├── docker-compose.yml                  # Multi-container orchestration
│
├── Backend/
│   ├── Dockerfile                      # Backend container setup
│   ├── env.example                     # Environment variables template
│   ├── package.json
│   ├── server.js                       # Backend entry point
│   ├── tests/
│   │   └── api.test.js                 # Jest + Supertest tests
│   │
│   └── src/
│       ├── app.js                      # Express app configuration
│       ├── config/                     # MongoDB configuration
│       ├── controllers/                # Route handlers
│       ├── middlewares/                # Auth + file middleware
│       ├── models/                     # Mongoose models
│       ├── routes/                     # Express routes
│       └── services/
│           ├── ai.service.js           # Gemini integration + PDF generation
│           └── rag.service.js          # Resume chunking, embedding, retrieval
│
├── Frontend/
│   ├── Dockerfile                      # Frontend container setup
│   ├── nginx.conf                      # Nginx production config
│   ├── package.json
│   ├── public/
│   │   └── _redirects                  # SPA routing fix
│   └── src/
│       ├── app.routes.jsx              # React routes
│       └── features/
│           ├── auth/                   # Authentication feature
│           └── interview/              # Interview dashboard and API services
│
└── README.md

---

🌐 Deployment

Backend on Render Web Service

Setting	Value
Root Directory	Backend
Build Command	npm install
Start Command	npm start

Environment Variables

MONGO_URI
JWT_SECRET
GOOGLE_GENAI_API_KEY
PINECONE_API_KEY
PINECONE_INDEX_NAME=gapwise-ai-rag-google
GOOGLE_EMBEDDING_MODEL=gemini-embedding-001
GOOGLE_EMBEDDING_DIMENSIONS=768
FRONTEND_ORIGIN=https://your-frontend-url.onrender.com

---

Frontend on Render Static Site

Setting	Value
Root Directory	Frontend
Build Command	npm install && npm run build
Publish Directory	dist

Environment Variables

VITE_API_URL=https://your-backend-url.onrender.com

---

Pinecone Setup

Create a plain dense Pinecone index for the RAG pipeline with:

Index name: gapwise-ai-rag-google
Vector type: Dense
Dimensions: 768
Metric: Cosine
Region: us-east-1

Notes:

• Do not use an integrated embedding index for this project.
• The backend generates embeddings itself using Google Gemini embeddings.
• The Pinecone index dimension must match GOOGLE_EMBEDDING_DIMENSIONS=768.

---

🧪 Running Tests

From Backend/

npm test

Uses

• Jest
• Supertest

Test location

Backend/tests/api.test.js

---

🔑 API Endpoints

Auth Routes

Method	Endpoint	Description
POST	/api/auth/register	Register user
POST	/api/auth/login	Login user
GET	/api/auth/logout	Logout user
GET	/api/auth/get-me	Current user

---

Interview Routes

Method	Endpoint	Description
POST	/api/interview/	Generate interview report
GET	/api/interview/	Fetch all reports
GET	/api/interview/report/:id	Fetch single report
POST	/api/interview/resume/pdf/:id	Generate and download resume PDF

---

🧠 AI and RAG Workflow

1. User submits job description and either resume PDF, self-description, or both
2. Resume PDF is parsed to text
3. RAG service chunks resume text, generates local embeddings, and stores a per-user collection
4. Job description is embedded and used to retrieve top relevant resume chunks
5. Retrieved chunks are sent to Gemini for structured report generation
6. Match score is normalized to 0 to 100 before persistence and UI rendering
7. Interview report is saved with ragMetadata fields used and chunksRetrieved
8. For resume download, Gemini generates resume HTML, then Puppeteer renders ATS-friendly PDF
9. PDF generation falls back to local Chrome/Edge binaries when needed for reliability

---

<<<<<<< HEAD MIT © Akshay Anand

4971bbfe600b7393b581d5e744459f64c6a1cf07