🛡️ Valora — Sensitive Data Guard

Valora is a powerful Data Loss Prevention (DLP) solution designed to protect sensitive information from leaking into third-party AI platforms like ChatGPT, Google Gemini, and Claude.

Built for Everyone: Whether you are an enterprise seeking to enforce corporate data policies or an individual user looking to protect your personal privacy, Valora operates through a real-time Chrome extension that actively monitors user input. It identifies and auto-masks generic Personally Identifiable Information (PII) alongside custom keywords before the data ever leaves the browser.

🚀 Key Features

• Real-time AI Platform Monitoring: Silently runs on AI interfaces (ChatGPT, Claude, Gemini) scanning for violations on the fly.
• Dual Modes of Operation:
  • Enterprise Mode: Multi-tenant architecture for organizations. Configure custom domains, keywords, and rule bypasses via an Admin Dashboard. Features automated auto-masking for company secrets.
  • Individual Mode: Perfect for personal safety. Protects everyday users from accidentally pasting credit cards, SSNs, API keys, or personal emails into AI models. Includes a generous free tier (up to 50 scans).
• Smart Auto-Masking: Automatically replaces custom company keywords with inline asterisks (***), ensuring context is retained while protecting sensitive variables.
• General PII Protection: Built-in detection for standard sensitive data (Credit Cards, SSNs, API Keys, Emails, Phone Numbers).
• Admin Dashboard Analytics (Enterprise): A modern, dark-mode React Dashboard featuring rich charts, timeline views, Top Offender lists, and platform usage metrics.
• Active User Heartbeats: Tracks active employee usage daily. Includes an automated chron job that emails administrators and users if they disable or abandon the protection extension.

📂 Project Structure

This repository contains the complete Valora ecosystem, divided into three main applications:

1. valora-admin/ (Frontend Dashboard)

   • A responsive, beautifully crafted Vite + React frontend dashboard.
   • Allows organization admins to track violations, manage rules, invite employees, and view rich data visualizations (Recharts).

2. valora-backend/ (Node.js API)

   • An Express & MongoDB backend handling authentication, rule delivery, and logging.
   • Multi-tenancy enabled via organizational scoping (orgId).
   • Includes automated cron jobs (inactivityChecker.js) using Nodemailer to alert inactive employees.

3. valora-extension-v2/ (Chrome Extension)

   • A Manifest V3 Chrome Extension.
   • Communicates seamlessly with the backend via background.js and uses a localized algorithm in content.js/detector.js to execute real-time regex matching and UI interceptions (blocking the send button, showing warning modals).

🛠️ Tech Stack

• Frontend: React, Vite, CSS Modules, Recharts
• Backend: Node.js, Express, MongoDB, Mongoose, JWT Authentication
• Extension: Vanilla JavaScript, Chrome Extension API (Manifest V3)
• Utilities: Nodemailer (SMTP alerting), Dotenv

🚦 Getting Started

Prerequisites

• Node.js (v18+ recommended)
• MongoDB Database (Local or MongoDB Atlas)

1. Setup the Backend

cd valora-backend
npm install

Create a .env file in the valora-backend directory:

MONGO_URI=mongodb://127.0.0.1:27017/valora
PORT=5000
JWT_SECRET=your_jwt_secret
SMTP_HOST=smtp.gmail.com
SMTP_PORT=587
SMTP_USER=your_email@gmail.com
SMTP_PASS=your_app_password
SMTP_FROM=your_email@gmail.com
INACTIVITY_DAYS=3

Run the server:

npm run dev

2. Setup the Admin Dashboard

cd valora-admin
npm install

Run the development server:

npm run dev

3. Load the Chrome Extension

1. Open Google Chrome and navigate to chrome://extensions/
2. Enable "Developer mode" in the top right corner.
3. Click "Load unpacked" and select the valora-extension-v2 folder.
4. Click the Valora puzzle piece icon in your browser to sign in to your Employee or Admin account.

🔒 Security & Privacy First

Valora's detection runs entirely inside the user's browser via the extension's content scripts. Keystrokes are never logged to the server. The backend only receives metadata when a violation policy is breached (e.g., the platform used and the type of data matched), ensuring maximum user privacy.

---

Developed for robust data compliance and enterprise AI safety.