AI Resume Screening System

This project is an AI-powered resume screening system developed as part of Task 3 of the Machine Learning Internship program.

The application analyzes uploaded PDF resumes and compares them against a job description to evaluate candidate suitability based on skill matching and text similarity.

Features

• Upload resume in PDF format
• Extract text from resumes automatically
• Compare resume with job description
• Generate resume match score
• Identify matching skills
• Detect missing skills
• Interactive web interface using Streamlit

Technologies Used

• Python
• Streamlit
• Scikit-learn
• TF-IDF Vectorization
• Cosine Similarity
• PyPDF2

Project Structure

FUTURE_ML_03/
│
├── data/
├── outputs/
├── src/
│   ├── task3.py
│   └── app.py
│
├── README.md
└── .gitignore

How the System Works

1. User uploads a resume PDF
2. Resume text is extracted using PyPDF2
3. Text is cleaned and processed
4. TF-IDF converts text into numerical vectors
5. Cosine Similarity calculates resume-job match score
6. Skills are extracted and compared
7. Matching and missing skills are displayed

Running the Project

1. Activate environment

source ../FUTURE_ML_01/ml_env/bin/activate

2. Install dependencies

pip install streamlit PyPDF2 scikit-learn pandas

3. Run the application

streamlit run src/app.py

Example

Job Description

Looking for a Python developer with Machine Learning, SQL, NLP, communication skills, and data analysis experience.

Output

• Resume Match Score
• Matching Skills
• Missing Skills

Future Improvements

• Multi-resume ranking system
• Better NLP models
• Resume keyword highlighting
• Online deployment
• Deep learning-based resume analysis

Author

Samarpit Gujral