🤖 Resume PII Masking & ATS Optimizer

A professional-grade NLP pipeline to automatically detect and mask Personally Identifiable Information (PII) in resumes and evaluate resume quality based on Applicant Tracking System (ATS) scoring. Built using the Hugging Face Transformers ecosystem and fine-tuned with custom data, this project simulates real-world applications of Natural Language Processing in HR tech and recruitment automation systems.

Key Features

Feature	Description
PII Masking	Detects and masks names, emails, phone numbers, and addresses using NER.
Resume Parsing	Handles large resumes (up to 2000+ words) with tokenizer support.
ATS Resume Optimization	Scores resumes based on keyword density, formatting, and clarity.
Job Description Matching	Optional feature to match resumes with specific job descriptions.
Hugging Face Integration	Fine-tune and deploy models directly on Hugging Face Hub.
Modular Architecture	Well-organized, scalable, and production-ready codebase.

📁 Folder Structure

resume_ats_project/
├── data/                  # Contains resume samples and PII-labeled training data
│   ├── resumes.json
│   └── pii_train.json
├── models/                # Directory to save fine-tuned models
│   └── ats_model/
├── resume_parser.py       # Tokenization, segmentation, and formatting
├── pii_trainer.py         # Script to fine-tune NER model
├── optimizer.py           # ATS scoring logic
├── infer.py               # Combines parsing, masking, and optimization
├── app.py                 # (Optional) Flask or Gradio interface
├── requirements.txt
└── README.md


---
Installation 
git clone https://github.com/your-username/resume-ats-optimizer.git
cd resume_ats_optimizer
pip install -r requirements.txt

---
Real-World Applications
This project mimics systems used by:
LinkedIn Talent Solutions (Resume scoring + redaction)
Amazon HR Automation (Internal resume screening tools)
Google Cloud AutoML NER for internal document pipelines
Infosys & TCS resume filtering portals

---
You can adapt it to:
Job matching portals
Candidate anonymization systems
Large-scale recruitment automation tools

---
License
Licensed under the MIT License.

---
Author
Karthikeyan M C
[email protected]