An AI-powered microservice application that automates resume–job matching. The system intelligently parses resumes, extracts skills using Natural Language Processing (NLP), and generates dynamic match scores for job descriptions — streamlining recruitment with accuracy and speed.
🔴 Live Demo: Link to the Deployed App
- AI-Powered Skill Extraction: Uses a dedicated Python + Flask + spaCy NLP microservice to identify core technical and soft skills from PDF resumes and job descriptions.
- Dynamic Match Scoring: Computes a real-time percentage match between a candidate’s resume and job description, with skill-gap analysis.
- Secure JWT Authentication: Includes complete user authentication and protected APIs using JSON Web Tokens.
- Full-Stack Microservice Architecture: Built with a React frontend, Node.js backend, and Python NLP service, deployed independently for scalability.
- Cloud-Native Deployment: Frontend hosted on Vercel, backend and NLP services on Render, and database on MongoDB Atlas.
- Modern, Responsive UI: Intuitive dark-themed interface built with React and Vite for smooth performance.
| Component | Technology |
|---|---|
| Frontend |   |
| Backend API |   |
| NLP Service |    |
| Database |  |
| Deployment |   |
-
React Frontend (Vercel): Handles authentication, resume uploads, and displays skill match scores.
-
Node.js Backend (Render): Manages users, JWT-based sessions, and routes NLP requests.
-
Python NLP Service (Render): Uses spaCy to extract relevant skills and return structured data to the backend.
git clone https://github.com/Yash-Jain24/AI-Resume-Matcher.git
cd AI-Resume-Matchercd nlp-microservice
pip install -r requirements.txt
python app.py
# Runs on http://localhost:5002cd ../backend
npm install
# Create a .env file
npm start
# Runs on http://localhost:5001cd ../frontend
npm install
npm run dev
# Runs on http://localhost:5173MONGO_URI=your_mongodb_atlas_connection_string
JWT_SECRET=your_super_secret_token_string
SPACY_SERVICE_URL=http://localhost:5002VITE_API_URL=http://localhost:5001/api- Frontend → Deployed on Vercel directly from
/frontend. - Backend & NLP Service → Deployed as independent Render web services.
- Database → Hosted on MongoDB Atlas (Free Tier Cluster).
Each service follows its own CI/CD pipeline for seamless updates and scalability.
Contributions and feature suggestions are welcome — open an issue or pull request.
Licensed under the MIT License.