Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Local-first privacy guard: anonymize your documents before sharing with LLMs.
| Date | Stars |
|---|---|
| 2026-07-31 | 419 |
| 2026-08-06 | 419 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
<div align="center"> <img src="report/images/mascot_shield.png" alt="rizzo-pii mascot — a purple hedgehog guarding a document with a shield" width="180" /> # rizzo-pii ### Local, reversible PII anonymization for Italian legal text **_Use frontier models without giving up your data._** <p> <img src="https://img.shields.io/badge/100%25-LOCAL-7c3aed?style=for-the-badge" alt="100% local" /> <img src="https://img.shields.io/badge/GDPR-BY%20DESIGN-7c3aed?style=for-the-badge" alt="GDPR by design" /> <img src="https://img.shields.io/badge/EU%20AI%20ACT-ALIGNED-7c3aed?style=for-the-badge" alt="EU AI Act aligned" /> </p> <p> <img src="https://img.shields.io/badge/params-%E2%89%880.3B-blue" alt="0.3B parameters" /> <img src="https://img.shields.io/badge/RAM-~0.5%20GB%20·%20CPU-blue" alt="0.5 GB RAM CPU" /> <img src="https://img.shields.io/badge/PII%20categories-22-blue" alt="22 PII categories" /> <img src="https://img.shields.io/badge/micro--F1-0.989-brightgreen" alt="0.989 micro-F1" /> <img src="https://img.shields.io/badge/offline-no%20API%20key-brightgreen" alt="offline, no API key" /> </p> **📄 [Read the full technical report (PDF)](report/rizzo-pii-report.pdf)** — model, dataset, method and experiments in detail <p> <a href="https://github.com/Rizzo-AI-Academy/rizzo-pii/releases/latest"><img src="https://img.shields.io/badge/⬇%20Download%20for-Windows-0078D6?style=for-the-badge&logo=windows&logoColor=white" alt="Download for Windows" /></a> <a href="https://github.com/Rizzo-AI-Academy/rizzo-pii/releases/latest"><img src="https://img.shields.io/badge/⬇%20Download%20for-macOS-000000?style=for-the-badge&logo=apple&logoColor=white" alt="Download for macOS" /></a> <a href="https://github.com/Rizzo-AI-Academy/rizzo-pii/releases/latest"><img src="https://img.shields.io/badge/⬇%20Download%20for-Linux-FCC624?style=for-the-badge&logo=linux&logoColor=black" alt="Download for Linux" /></a> </p> <sub>🪟 **Windows installer** · 🍎 **macOS** (Apple Silicon) · 🐧 **Linux AppImage** — all available now</sub> </div> **`rizzo-pii:0.3B`** is a lightweight, CPU-friendly, **Italian-first** token-classification model (≈0.3B parameters, [mmBERT](https://huggingface.co/jhu-clsp/mmBERT-base) / ModernBERT backbone) that detects **22 categories of personal data** — including the Italian-legal identifiers (**codice fiscale**, **partita IVA**, **dati catastali**) that no other open model covers — and drives a fully **reversible** anonymization workflow: <div align="center"> 🔒 **anonymize locally** → 🏷️ **placeholder + reversible local dictionary** → ☁️ **frontier LLM** → 🔓 **restore locally** </div> It is built for law firms, accountants, notaries and anyone bound by the **GDPR** who wants to keep using ChatGPT / Claude / Gemini on sensitive documents **without ever sending the real data out**. <table> <tr> <td width="120" align="center" valign="middle"> <img src="report/images/mascot_doc.png" alt="rizzo-pii redacting a document" width="110" /> </td> <td valign="middle"> | ≈0.3B | ~0.5 GB | 22 | 0.989 | |:---:|:---:|:---:|:---:| | parameters (mmBERT-base) | RAM footprint, CPU | PII categories | micro-F1 (real IT validation) | The hedgehog mascot does one job: it grabs your document, blacks out every identifier, and **stays inside the EU** while doing it. </td> </tr> </table> --- ## The problem: convenience is leaking your data People summarize contracts, draft replies and ask legal questions simply by **pasting the document in**. It is fast and useful — and it quietly moves enormous amounts of personal and confidential data off the user's device. Names, addresses, tax codes, IBANs, health details, unsigned-contract clauses: all of it crosses the network to servers the user does not control, where it may be logged, cached, retained or exposed in a breach. For a law firm or a hospital this is not hypothetical; under the GDPR it can be a direct compliance failure. The intuitive fix is to **stop sending data out** and run an open model loca
Excerpt of 29,646 characters
Read on GitHubWould you bet a product on this? Bounded 0–100 and slow moving.
matched fp:2ac9b46f1cf0cf96, llm:Repository description: 'Local-first privacy guard: anonymize your documents before sharing with LLMs.' Topics include anonymization, gdpr, named-entity-recognition, pii, token-classification, transformers, llm, privacy.
matched fp:2ac9b46f1cf0cf96, llm:Repository description: 'Local-first privacy guard: anonymize your documents before sharing with LLMs.' Topics include anonymization, gdpr, named-entity-recognition, pii, token-classification, transformers, llm, privacy.
matched fp:2ac9b46f1cf0cf96, llm:Repository description: 'Local-first privacy guard: anonymize your documents before sharing with LLMs.' Topics include anonymization, gdpr, named-entity-recognition, pii, token-classification, transformers, llm, privacy.