Show HN: A 0.3B model that redacts PII in all 24 EU languages offline

Category: ai-ml

Tags: pii-detection, gdpr-compliance, multilingual-nlp

Score: 7.7/10 (Innovation: 7, Technical: 8, Documentation: 8, Utility: 8)

This project is a multilingual PII detection model trained natively on all 24 EU languages, covering 36 entity classes including GDPR Article 9 special categories often missed by other tools. It provides quantized ONNX weights for CPU-friendly production use, making it a practical solution for compliance-driven redaction workflows in European contexts.

Target audience: compliance engineers, privacy engineers, backend devs

Repository: https://huggingface.co/bardsai/eu-pii-anonimization-multilang

View on Hacker News