A powerful, AI-powered PII detection and redaction tool built with SwiftUI

On-device Model

The machine learning model for this project is available at: https://github.com/oadultradeepfield/redactkit-on-device-model

🚀 Overview

RedactKit is an intelligent text processing application that automatically detects and redacts Personally Identifiable Information (PII) from text content. Using advanced machine learning models and Natural Language Processing, it identifies sensitive information like emails, phone numbers, addresses, SSNs, credit card numbers, and more.

✨ Key Features

• 🧠 AI-Powered Detection: Uses CoreML and NeuroBERT for accurate PII identification
• 🎨 Beautiful UI: Modern SwiftUI interface with stunning animations and effects
• ⚡ Real-time Processing: Instant text analysis and redaction
• 📚 History Management: Persistent storage of processed content using SwiftData
• 🔄 Animated Redaction: Smooth animations showing the redaction process
• 📋 Clipboard Integration: Easy copy/paste functionality

🔍 Supported PII Types

RedactKit can detect and redact the following types of sensitive information:

Type	Examples	Description
Person Names	John Doe, Sarah Smith	Individual names and identities
Email Addresses	user@example.com	Email addresses and contact information
Phone Numbers	(555) 123-4567, +1-800-555-0199	Various phone number formats
Physical Addresses	123 Main St, Anytown, CA 90210	Street addresses and locations
Social Security Numbers	123-45-6789	SSNs and similar identifiers
Credit Card Numbers	4111-1111-1111-1111	Payment card information
Dates	01/15/1990, Jan 15, 1990	Birth dates and other sensitive dates

🛠️ Technology Stack

• Framework: SwiftUI 5.0+
• Language: Swift 5.9+
• ML Framework: CoreML 5.0+
• NLP: Natural Language Framework
• Tokenization: HuggingFace Tokenizers (NeuroBERT-Mini)
• Data Persistence: SwiftData
• Animation: SwiftUI Animations with Matched Geometry Effects
• Minimum Target: macOS 14.0+

📦 Installation

Prerequisites

• macOS 14.0 or later
• Xcode 15.0 or later
• Swift 5.9 or later

Building from Source

1. Clone the repository

   git clone https://github.com/Spchdt/RedactKit.git
   cd RedactKit

2. Open in Xcode

   open RedactKit.xcodeproj

3. Build and Run

   • Select your target device/simulator
   • Press Cmd + R to build and run

Dependencies

The project includes:

• CoreML model (PIIDetectionModel.mlpackage)
• NeuroBERT-Mini tokenizer

🎨 Design Philosophy

RedactKit emphasizes:

• User Experience: Intuitive interface with beautiful animations
• Performance: Optimized ML inference for real-time processing
• Security: Safe handling of sensitive information
• Accessibility: Clean, readable design with clear visual feedback