Fidelius

This project is a data masking and obfuscation tool designed to enhance privacy and security by intelligently handling Personally Identifiable Information (PII).

Description

Fidelius is a localized data masking tool utilizing Generative AI. It allows users to upload files and select between two methods of data protection: redaction or obfuscation. The tool is currently compatible with two types of files: CSV/Excel and PDF/Text documents. It is designed to protect sensitive information while maintaining the integrity and usability of the data.

Demo Video

How To Use in NVIDIA AI Workbench

1. Clone this repository:

   git clone https://github.com/Sreehari78/Fidelius.git

2. Set Up the Application in Workbench:

   • Name: Fidelius

   • Class: Web Application

   • Icon URL: Leave blank

   • Start Command:

     python run.py

   • Port: 5000 (ensure it's not used by another application)

   • Health Check Command:

     curl -f "http://localhost:5000"

   • Stop Command:

     pkill -f "python run.py"

   • Auto Launch: Selected

   • URL: http://localhost:5000

   • URL Command: Leave blank

   • User Message: Leave blank

3. Run the Application:

   Once you have configured the application in the NVIDIA AI Workbench, click start in your Workbench environment to run Fidelius. If running as an app doesn't work, open in VS Code and navigate to /code and run run.py with python3 run.py or navigate to code/client and start the client with npm run dev. Navigate to code/server and start the server with python3 server.py.

4. Upload your files: The tool currently supports CSV/Excel and PDF/Text formats.

5. Select the desired operation: Choose to obfuscate or redact the identified PII.

6. Generate the new file: The system will process your file and provide a new version with the masked data.

Features

• Data Masking: Securely masks sensitive data using advanced AI techniques.
• File Support: Handles CSV/Excel and PDF/Text files effectively.
• User Customization: Allows users to specify additional PII fields for masking according to their needs.
• Local Processing: Ensures that all data processing is done locally, enhancing security.

Future Enhancements

• Support for more file types and formats.
• Improved AI models for more accurate PII detection and masking.
• Enhanced user interface for easier interaction and customization.

---

This README provides a clear and structured explanation of the Fidelius project, its functionality, usage instructions, and future development plans, making it easy for users to understand and start using the tool.