: RESUMEAI :

🤖 AI Resume Analyser Web Application:

A portfolio-grade full-stack application that ingests a PDF resume and a job description, runs them through an AI pipeline (Google Gemini 1.5 Flash + embeddings), and returns a match score, skill gap analysis, AI-powered insights, and an auto-improved resume with diff view.

---

✨ Features

Feature	Description
Match Score	Composite score: 60% semantic (embeddings) + 40% keyword coverage
Skill Gap Analysis	Exact skills you're missing vs. skills you have
Radar Chart	Visual comparison across 5 dimensions
AI Insights	Gemini-powered strengths, suggestions, assessment
Resume Rewrite	ATS-optimised rewrite tailored to the job
Diff View	Side-by-side word-level diff of original vs improved

---

🏗️ Architecture

frontend/   → React + Vite (port 5173)
backend/    → Node.js + Express (port 5000)
PostgreSQL  → Optional (pgvector for embeddings) — in-memory fallback if not set
Gemini API  → Gemini 1.5 Flash + text-embedding-004

---

🚀 Quick Start

Prerequisites

• Node.js 18+
• A Google Gemini API key
• (Optional) PostgreSQL with pgvector

1. Clone

git clone <repo-url>
cd ai-resume-analyser

2. Backend Setup

cd backend
cp .env.example .env
# Edit .env — add your GEMINI_API_KEY
npm install
npm run dev
# Server starts on http://localhost:5000

3. Frontend Setup

cd frontend
npm install
npm run dev
# App opens on http://localhost:5173

4. Optional: PostgreSQL + pgvector

If you want persistent storage:

1. Create a PostgreSQL database (local or Neon)
2. Install pgvector: CREATE EXTENSION vector;
3. Run the schema: psql $DATABASE_URL -f backend/src/db/schema.sql
4. Set DATABASE_URL in backend/.env

---

🌍 Environment Variables

backend/.env

Variable	Required	Default	Description
GEMINI_API_KEY	✅ Yes	—	Your Gemini API key
GEMINI_MODEL	No	gemini-1.5-flash	Generative model
GEMINI_EMBEDDING_MODEL	No	text-embedding-004	Embedding model
PORT	No	5000	Backend port
CLIENT_ORIGIN	No	http://localhost:5173	Frontend URL (for CORS)
DATABASE_URL	No	—	PostgreSQL connection string

---

📡 API Reference

Method	Endpoint	Description
GET	/health	Health check
POST	/api/resume/upload	Upload PDF resume (multipart/form-data)
POST	/api/resume/improve	Generate improved resume
POST	/api/job/submit	Submit job description text
POST	/api/analysis/match	Run match analysis

---

🗂️ Project Structure

ai-resume-analyser/
├── backend/
│   ├── src/
│   │   ├── app.js              # Express app
│   │   ├── server.js           # Entry point
│   │   ├── config/env.js       # Env config
│   │   ├── routes/             # API routes
│   │   ├── controllers/        # Request handlers
│   │   ├── services/
│   │   │   ├── pdfParser.js    # PDF text extraction
│   │   │   ├── llm.js          # OpenAI prompts
│   │   │   ├── embeddings.js   # Embedding generation
│   │   │   └── matcher.js      # Cosine similarity + scoring
│   │   ├── db/
│   │   │   ├── index.js        # PostgreSQL pool
│   │   │   └── schema.sql      # DB schema (pgvector)
│   │   └── utils/             # Logger, errorHandler
│   └── package.json
├── frontend/
│   ├── src/
│   │   ├── pages/              # Home, Analyze, Results
│   │   ├── components/         # UploadResume, JobInput, etc.
│   │   ├── hooks/useApi.js     # Axios hook
│   │   └── utils/formatters.js # Score utilities
│   └── package.json
└── README.md

---

🚢 Deployment

Backend → Render

1. Connect repo → New Web Service
2. Root directory: backend
3. Start command: npm start
4. Add environment variables in Render dashboard

Frontend → Vercel

1. Connect repo → New Project
2. Root directory: frontend
3. Build command: npm run build
4. Set VITE_API_URL to your Render backend URL

Database → Neon

1. Create a Neon project (free tier available)
2. Enable pgvector: CREATE EXTENSION vector;
3. Run backend/src/db/schema.sql
4. Set DATABASE_URL in Render env vars

---

🛠️ Tech Stack

Layer	Technology
Frontend	React 18, Vite, React Router, Recharts, react-diff-viewer
Backend	Node.js, Express, Multer, pdf-parse
AI	Google Gemini 1.5 Flash, text-embedding-004
Database	PostgreSQL + pgvector (optional)
Styling	Vanilla CSS (design system with CSS variables)

---

📄 License

MIT