ATS Resume Checker (MERN)

AI-powered ATS Resume Checker built with React + Express + MongoDB + Gemini.

Architecture

+--------------------+        HTTP (REST)         +-------------------------+
| React (Vite)       |  <---------------------->  | Node.js + Express API   |
| frontend           |                            | backend/server.js       |
+--------------------+                            +-----------+-------------+
                                                              |
                                               Mongoose       |
                                                              v
                                                    +-------------------+
                                                    | MongoDB Atlas     |
                                                    | resume analyses   |
                                                    +-------------------+
                                                              |
                                                              | SDK call
                                                              v
                                                    +-------------------+
                                                    | Google Gemini API |
                                                    +-------------------+

Backend Structure

backend/
├── server.js
├── package.json
├── .env
├── config/
│   └── db.js
├── models/
│   └── ResumeAnalysis.js
├── routes/
│   └── resume.routes.js
├── controllers/
│   └── resume.controller.js
├── services/
│   └── resume.service.js
└── middleware/
    ├── upload.middleware.js
    └── error.middleware.js

Features

• Upload resume in PDF / DOCX / TXT (up to 10MB)
• ATS analysis against optional job description
• Keyword match and missing-keyword detection
• ATS score + feedback + suggestions
• ATS-optimized resume generation
• Resume analysis history with server-side limit support
• Automatic in-memory fallback when MongoDB is unavailable

API Endpoints

Base URL: http://localhost:5000/api/resume

• POST /analyze
  • multipart/form-data
  • fields:
    • file (required)
    • jobDescription (optional)
• POST /generate-ats
  • JSON body:

    {
      "resumeText": "...",
      "jobDescription": "..."
    }

• GET /history
  • query params:
    • limit (optional, 1-100, default 25)
    • includeText (optional, 1/true/yes to include extracted resume text)
• GET /health
  • returns service status, uptime, storage mode, and MongoDB connection state

Mongoose Schema

const ResumeAnalysisSchema = new mongoose.Schema({
  filename: String,
  resumeText: String,
  jobDescription: String,
  atsScore: Number,
  matchedKeywords: [String],
  missingKeywords: [String],
  feedback: String,
  suggestions: [String],
  optimizedResume: String,
  createdAt: { type: Date, default: Date.now }
});

Environment Variables

Backend (backend/.env)

PORT=5000
MONGODB_URI=your_mongodb_atlas_connection_string
GEMINI_API_KEY=your_gemini_api_key
GEMINI_MODEL=gemini-1.5-flash

Frontend (frontend/.env)

VITE_API_URL=http://localhost:5000
VITE_ADMIN_EMAIL=
VITE_API_TIMEOUT_MS=20000

VITE_ADMIN_EMAIL is optional and only used to show/hide the admin dashboard link. VITE_API_TIMEOUT_MS is optional and controls frontend API timeout in milliseconds. If VITE_ADMIN_EMAIL is not set, the first registered account becomes the local owner/admin by default.

Run Locally

1. Backend

cd backend
npm install
npm run dev

Backend runs on http://localhost:5000.

2. Frontend

cd frontend
npm install
npm run dev

Frontend runs on http://localhost:5173.

Quick Checks

Backend health

curl http://localhost:5000/health

History (latest 10, without full resume text)

curl "http://localhost:5000/api/resume/history?limit=10"

Frontend Routes

• /upload -> upload resume + analyze
• /result -> ATS score, keywords, suggestions, optimized resume download
• /history -> past analysis records from MongoDB
• /admin -> admin dashboard (owner/admin only)

Tech Stack

• Frontend: React, Vite
• Backend: Node.js, Express
• Database: MongoDB Atlas + Mongoose
• File Parsing: pdf-parse, mammoth, multer
• AI: @google/generative-ai (Gemini)