Resume Matcher is an open-source AI tool designed to help job seekers tailor their resumes to specific job descriptions. It analyzes both documents to identify relevant keywords, improve readability, and provide insights for ATS compatibility, aiming to increase job application success rates.

How It Works

The tool parses resumes and job descriptions using Python, mimicking Applicant Tracking System (ATS) functionality. It extracts key terms and keywords from job descriptions using machine learning and identifies key themes. Vector similarity, powered by FastEmbed, is then used to quantify how closely a resume matches the job description, indicating its ATS-friendliness.

Quick Start & Requirements

• Install dependencies: pip install -r requirements.txt
• Prepare data: Place PDF resumes in Data/Resumes and job descriptions in Data/JobDescription.
• Parse resumes: python run_first.py
• Run application: streamlit run streamlit_app.py
• Requires Python 3.10+.
• Docker support is available.
• Pre-computation of vector similarity is done for performance; GPU/significant RAM is recommended for local computation.
• See How to Install for detailed setup.

Highlighted Details

• Utilizes FastEmbed for efficient vector similarity calculations.
• Provides keyword extraction and key term identification from job descriptions.
• Offers a demo at https://resume-matcher.streamlit.app/.
• Includes a full-stack web application (Next.js, React, FastAPI) with a mocked frontend for demonstration.

Maintenance & Community

• Active community with Discord server (https://dsc.gg/resume-matcher) and Twitter presence (https://twitter.com/srbhr).
• Encourages contributions via Pull Requests and Issues.
• Development is supported by donations via Buy Me a Coffee and GitHub Sponsors.

Licensing & Compatibility

• Licensed under Apache 2.0.
• Compatible with commercial use and closed-source linking.

Limitations & Caveats

The full-stack web application's results are currently mocked and not based on real data processing. Local computation of vector similarity may require significant GPU and RAM resources.