Show HN: Rampart on CoreML
A CoreML implementation of National Design Studio's Rampart PII detection model has been released as a Swift package, enabling on-device personally identifiable information detection in iOS/macOS apps.
Background
- Rampart is an open-source model by National Design Studio that detects Personally Identifiable Information (PII) — things like Social Security numbers, credit card numbers, phone numbers, and email addresses — in text. It's designed to run locally rather than calling a cloud API.
- CoreML is Apple's machine learning framework for running models on iPhones, iPads, and Macs without sending data to a server. This project ports the Rampart model to CoreML, which means PII detection can happen entirely on-device on Apple hardware.
- The significance: many companies use cloud services for PII redaction, which requires sending sensitive text to an external server. On-device detection keeps the data private. This packaging makes it easy for iOS/macOS developers to add real-time PII scanning to their apps via a Swift package.
- The author provides both the converted CoreML model and a Swift package for straightforward integration.