deformatic

deformatic / OPENAI-Privacy-Filter-Reversible-Tokenization

Public

OPENAI Privacy Filter with a reversible tokenization vault layer

20
7
100% credibility
Found May 08, 2026 at 20 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

This repository adds reversible pseudonymization to OpenAI's local Privacy Filter model, replacing detected PII spans with stable indexed placeholders and storing originals in a protected vault for later restoration.

How It Works

1
🔍 Discover the Privacy Tool

You hear about a helpful tool that spots private info like names and emails in your text and hides them safely while keeping track of who is who.

2
📦 Set It Up Easily

You add the tool to your computer with a simple command, and it grabs the needed brainpower from a trusted place.

3
✨ Protect Your Text

You give it a sentence with personal details, and it swaps them out for numbered stand-ins like Person 1, creating a secret list to unlock later.

4
đź”’ See the Magic

Your text now looks clean and safe to share, but you hold the key to bring back the real details anytime you need.

5
🔓 Unlock When Ready

Use the secret list to swap back and get your original text perfectly restored.

âś… Privacy Perfected

Now you can work with sensitive info confidently, sharing safe versions while keeping everything recoverable and connected.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 20 to 20 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is OPENAI-Privacy-Filter-Reversible-Tokenization?

This Python library extends OpenAI's local privacy filter with a reversible tokenization layer, detecting PII like names, emails, and secrets in text, then swapping them for stable placeholders such as while storing originals in a vault file for later restoration. It solves the gap between irreversible redaction—which breaks entity relationships in pipelines—and full PII exposure, enabling pseudonymized text for downstream LLMs or services. Users get CLI commands like `opf --recoverable --vault-out vault.json` for batch processing and a simple API for tokenizing/restoring in openai github python workflows.

Why is it gaining traction?

Unlike basic openai privacy filter tools that permanently erase data, this adds consistent tokens across documents via a shared vault, preserving context for tasks like LLM enrichment or review. It's backward-compatible, model-agnostic, and production-ready with vault encryption guidance, making it a smart openai github alternative for privacy-aware pipelines without vendor APIs. Developers hook on the audit-friendly JSON vaults and cross-input stability for openai privacy api integrations.

Who should use this?

Data engineers piping text through openai github actions or copilot-like tools needing PII safeguards. Compliance teams at startups handling user data in openai github apps, wanting reversible flows before openai privacy policy updates. ML ops folks building local openai privacy filter layers for batch jobs with whisper or codex outputs.

Verdict

Grab it for prototyping reversible PII handling—solid docs, CLI, and tests make it dev-friendly despite 19 stars and 1.0% credibility signaling early maturity. Scale cautiously in prod until more adoption; pair with KMS for vaults.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.