Data Anonymization Service

SDV based tabular data anonymization tool with a 4-step workflow:

1. Upload data (CSV / XLS / XLSX)
2. Detect PII
3. Configure rules (exempt_columns / force_pii_columns)
4. Generate, preview, compare, and download synthetic data in csv format

Project Structure

• src/
  • api.py - FastAPI endpoints
  • generate_metada.py - metadata generation and config save
  • Gaussian_Coupla_syntheticdata_generator.py - SDV synthetic data pipeline
• ui/ - frontend page and step scripts
• results/ - runtime outputs
  • uploaded_data/ - uploaded files converted to CSV
  • meta_data/ - generated metadata and user config
  • GaussianCopula_results/ - synthetic data, quality reports, scores
• raw_data/ - optional mounted input folder (Docker use)

Requirements

• Python 3.9+
• Dependencies in requirements-api.txt
• spaCy model:
  • python -m spacy download en_core_web_sm

Install through Docker

1) Prerequisites

• Install Docker Desktop
• Make sure Docker is running

2) Run

git clone https://github.com/gyt197/Data-Anonymization-Service.git
cd Data-Anonymization-Service
docker compose up --build

3) Access

App: http://localhost:8000

4) Stop

docker compose down

Install from Github

1) Clone the repository

git clone https://github.com/gyt197/Data-Anonymization-Service.git

2) Navigate into the project

cd Data-Anonymization-Service

3) Create a virtual environment

python -m venv venv

4) Activate the virtual environment

On Windows

my-env\Scripts\activate

5) Install dependencies

pip install -r requirements-api.txt
pip install sdv python-multipart
python -m spacy download en_core_web_sm

6) Run the server

uvicorn src.api:app --host 0.0.0.0 --port 8000 --reload

License

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