A FastAPI-based application designed to reduce hiring bias by parsing, anonymizing, and semantically matching resumes and job descriptions.

• Resume Parsing: Extract structured data from PDF resumes using pdfplumber.
• Anonymization: Remove personally identifiable information (PII) to ensure unbiased screening using spacy NER.
• Semantic Matching: Match candidates to jobs based on skills and context using sentence-transformers.
• REST API: Built with FastAPI for high performance and easy integration.
• React Frontend: Modern UI built with React and Vite.

• Python 3.9+ (Python 3.14 is currently unsupported by some dependencies)
• Node.js 18+ (for frontend)

1. Clone the repository

   git clone <repository_url>
   cd hackthebiasproject

2. Create a Virtual Environment

   python3.11 -m venv venv

3. Activate the Virtual Environment

   • macOS/Linux:

     source venv/bin/activate

   • Windows:

     .\venv\Scripts\activate

4. Install Dependencies

   pip install -r requirements.txt

5. Download spaCy Model

   python -m spacy download en_core_web_sm

6. Run the Backend Server

   uvicorn main:app --reload

   • API: http://127.0.0.1:8000
   • Docs: http://127.0.0.1:8000/docs

1. Navigate to frontend folder

   cd frontend

2. Install dependencies

   npm install

3. Run the development server

   npm run dev

   • Frontend: http://localhost:3000

hackthebiasproject/
├── app/
│   ├── routes/          # API endpoints
│   └── services/        # Business logic
├── core/
│   └── database.py      # SQLite database with SQLModel
├── models/              # Pydantic models
├── utils/
│   ├── anonymizer.py    # PII removal
│   ├── parser.py        # PDF extraction
│   └── semantics.py     # NLP matching
├── frontend/            # React + Vite frontend
├── main.py              # FastAPI entry point
└── requirements.txt     # Python dependencies