CV Embed: AI-Powered Resume Matching System

CV Embed is an AI resume analyzer that matches candidates with job descriptions using SBERT/GloVe/Doc2Vec models. Calculates compatibility scores, suggests top job matches, and generates real opportunities. Supports PDF/DOCX uploads with Flask backend and Gemini AI integration.

Key Features

• AI-Powered Matching: Semantic analysis using multiple NLP models
• Multi-Step Workflow: Intuitive 3-step process for resume and job analysis
• File & Text Support: Upload PDF/DOCX files or paste text directly
• Visual Results: Clear similarity score visualization with color coding
• Top Job Suggestions: Identifies best-fitting roles with similarity percentages
• Job Recommendations: Real job opportunities based on profile
• Privacy Focused: Automatic file cleanup after processing

---

Technology Stack

Component	Technology
Frontend	HTML5, CSS3, JavaScript
Backend	Python, Flask
AI Engine	Sentence Transformers, Gensim
Text Processing	NLTK, PyPDF2, python-docx
API Integration	Google Gemini API
Deployment	Docker (optional)

---

Project Structure

project_root/
├── model/
│   ├── inference/
│   ├── preprocess/
│   └── training/
├── trained_models/
├── templates/
│   ├── index.html
│   ├── result.html
│   └── upload.html
├── static/
├── data/
├── utils/
├── requirements.txt
├── config.py
├── data_loader.py
├── matching_engine.py
├── resume_processor.py
└── app.py

Website Interface

1. Resume Upload Page

• Users can upload or paste their resumes on this page.
  

2. job description Upload Page

• Users can upload or paste their resumes on this page.
  

3. Results Page

• This page displays the top 5 job descriptions that match the uploaded resume.
  

---

Installation & Setup

Prerequisites

• Python 3.8+
• pip package manager
• Google Gemini API key

Steps

1. Clone the repository

   git clone https://github.com/Samitha-Edirisinghe/AI-Powered-Resume-Matching-System-CVEmbed-.git
   cd cv-embed

2. Create a virtual environment

   # Windows
   python -m venv venv
   venv\Scripts\activate
   
   # macOS/Linux
   python3 -m venv venv
   source venv/bin/activate

3. Install dependencies

   pip install -r requirements.txt

4. Download NLTK data

   python -c "import nltk; nltk.download('stopwords'); nltk.download('punkt')"

5. Set environment variables
   Create a .env file in project root:

   GEMINI_API_KEY=your_api_key_here

6. Run the application

   python app.py

   Then open http://localhost:5000 in your browser.

---

Usage Workflow

1. Upload Resume

   • Select a PDF/DOCX file or paste resume text
   • Click Continue

2. Upload Job Description

   • Select a job description file or paste text
   • Choose matching algorithm (SBERT recommended)
   • Click Analyze

3. View Results

   • See your match percentage
   • Explore top job suggestions
   • Compare resume and job description
   • Browse real job opportunities
   • Click Analyze More to start over

---

AI Models

The system supports three matching algorithms:

1. SBERT (Recommended)

   • Uses Sentence Transformers for semantic similarity
   • Trained on job description–resume pairs

2. GloVe

   • Word embedding–based matching
   • Uses average word vectors

3. Doc2Vec

   • Document embedding approach
   • Trained on a resume corpus

---

Data Sources

Dataset	Source
job_cv_supervised	Hugging Face
resume-job-description-fit	Hugging Face
Resume-Dataset	Hugging Face
IT Job Post Descriptions	Kaggle

---

Job Taxonomy

The system matches resumes to these roles:

• System Administrator
• Database Administrator
• Web Developer
• Security Analyst
• Network Administrator
• Data Scientist
• DevOps Engineer
• Cloud Engineer
• Machine Learning Engineer
• Software Engineer

---

Configuration

Edit config.py to:

• Change active model
• Modify job taxonomy
• Add/remove datasets
• Adjust model parameters

---

Contributing

1. Fork the repository
2. Create your feature branch (git checkout -b feature/your-feature)
3. Commit your changes (git commit -m 'Add some feature')
4. Push to the branch (git push origin feature/your-feature)
5. Open a pull request

---

License

This project is licensed under the MIT License. See the LICENSE file for details.

---

Contact

For questions or feedback, reach out at:
randikanilupul123@gmail.com