🔒 Clyra - AI Privacy & Compliance Platform

Advanced PII Detection, GDPR Compliance, and EU AI Act Monitoring with MCP Integration

Clyra is a comprehensive privacy and compliance platform that combines cutting-edge AI technology with regulatory compliance tools. Built with modern web technologies and powered by lovable.dev, it provides real-time PII detection, anonymization, and compliance monitoring for GDPR and EU AI Act requirements.

🌟 Key Features

🔍 Advanced PII Detection: Multiple NER models with confidence scoring
🛡️ GDPR & AI Act Compliance: Real-time monitoring and risk assessment
🎤 Speech-to-Text Integration: OpenAI Whisper for audio transcription
🔗 MCP Server: Model Context Protocol integration for AI assistants
📊 Compliance Dashboard: Real-time monitoring and analytics
🌐 Modern Frontend: Built with React, TypeScript, and Tailwind CSS
⚡ FastAPI Backend: High-performance PII detection API
🗄️ Vector Database: Weaviate integration for data storage

🏗️ Architecture Overview

┌─────────────────────────────────────────────────────────────┐
│                        Clyra Ecosystem                      │
├─────────────────────────────────────────────────────────────┤
│  Frontend (React/Vite)     │  Backend (FastAPI)            │
│  ├─ Compliance Dashboard   │  ├─ PII Detection API        │
│  ├─ Real-time Monitoring   │  ├─ Hugging Face Integration │
│  ├─ Speech-to-Text UI      │  ├─ Multiple NER Models       │
│  └─ Admin Interface         │  └─ Anonymization Tools      │
├─────────────────────────────────────────────────────────────┤
│  MCP Server (Cloudflare)   │  Vector DB (Weaviate)         │
│  ├─ GitHub OAuth           │  ├─ User Login Tracking      │
│  ├─ PII Detection Tools     │  ├─ Compliance Data          │
│  ├─ Compliance Prompts      │  └─ Audit Logs              │
│  └─ AI Act Monitoring       │                              │
└─────────────────────────────────────────────────────────────┘

🚀 Quick Start

Prerequisites

Node.js 18+ and npm
Python 3.8+ (for PII detection API)
GitHub OAuth App (for MCP authentication)
Hugging Face API Token (for PII detection)
OpenAI API Key (for speech-to-text)

1. Clone and Install

git clone <your-repo-url>
cd clyra
npm install

2. Environment Setup

# Copy the comprehensive environment template
cp env.example .env

# Edit .env with your actual values
nano .env

Required Environment Variables:

# GitHub OAuth (Required)
GITHUB_CLIENT_ID=your_github_client_id
GITHUB_CLIENT_SECRET=your_github_client_secret

# Hugging Face API (Required for PII detection)
HF_API_TOKEN=your_huggingface_token

# OpenAI API (Required for speech-to-text)
OPENAI_API_KEY=your_openai_api_key

3. Start the Services

Terminal 1 - PII Detection API:

# Activate Python virtual environment
source presidio_env/bin/activate

# Start FastAPI server
cd streamlit
uvicorn hf_api_fastapi:app --reload --port 8000

Terminal 2 - Frontend:

cd frontend
npm run dev

Terminal 3 - MCP Server (Optional):

npm run dev

4. Access the Application

Frontend: http://localhost:5173
PII API: http://localhost:8000
API Docs: http://localhost:8000/docs

📁 Project Structure

clyra/
├── 📁 frontend/                 # React/Vite Frontend (Lovable.dev)
│   ├── 📁 src/
│   │   ├── 📁 components/        # UI Components
│   │   │   ├── Dashboard.tsx    # Compliance Dashboard
│   │   │   ├── AIInputField.tsx # Speech-to-Text Input
│   │   │   └── ui/              # shadcn/ui Components
│   │   ├── 📁 pages/            # Application Pages
│   │   ├── 📁 lib/              # Utilities & API Clients
│   │   └── App.tsx              # Main Application
│   └── package.json             # Frontend Dependencies
├── 📁 src/                      # MCP Server (Cloudflare Worker)
│   ├── index.ts                 # Main MCP Server
│   ├── pii-client.ts           # PII API Client
│   ├── weaviate.ts              # Vector DB Integration
│   └── auth/                    # GitHub OAuth Handlers
├── 📁 streamlit/                # FastAPI Backend
│   ├── hf_api_fastapi.py        # PII Detection API
│   ├── README_FASTAPI.md        # API Documentation
│   └── test_fastapi.py          # API Tests
├── 📁 presidio_env/             # Python Virtual Environment
├── 📄 env.example               # Environment Template
├── 📄 wrangler.jsonc            # Cloudflare Worker Config
├── 📄 mcp-inspector.json        # MCP Inspector Config
└── 📄 README.md                 # This File

🔧 Component Details

1. 🎨 Frontend (Lovable.dev Power)

Technology Stack:

React 18 with TypeScript
Vite for fast development
Tailwind CSS for styling
shadcn/ui for components
React Query for state management
React Router for navigation

Key Features:

Compliance Dashboard: Real-time monitoring of GDPR and AI Act compliance
Speech-to-Text: OpenAI Whisper integration for audio input
PII Detection: Real-time sensitive data detection
Admin Interface: Management and configuration tools
Responsive Design: Mobile-first approach

Built with Lovable.dev:

Modern, accessible UI components
Optimized performance
Type-safe development
Hot reloading and instant preview

2. ⚡ Backend (FastAPI)

Technology Stack:

FastAPI for high-performance API
Hugging Face Models for PII detection
Pydantic for data validation
Uvicorn for ASGI server

Available Models:

obi/deid_roberta_i2b2 - Medical de-identification (Recommended)
StanfordAIMI/stanford-deidentifier-base - Stanford de-identifier
dslim/bert-base-NER - General NER
Jean-Baptiste/roberta-large-ner-english - RoBERTa Large NER
dslim/bert-large-NER - BERT Large NER

API Endpoints:

GET /health - Health check and available models
GET /models - List available NER models
POST /detect - Detect PII entities
POST /anonymize - Anonymize PII with multiple operations

Anonymization Operations:

Redact: Remove PII completely
Replace: Replace with entity tags (<PERSON>, <EMAIL>)
Mask: Replace with asterisks (********)
Highlight: Mark PII for review

3. 🔗 MCP Server (Model Context Protocol)

Technology Stack:

Cloudflare Workers for serverless deployment
TypeScript for type safety
Hono for lightweight web framework
GitHub OAuth for authentication

MCP Tools:

hello - Basic greeting tool
pii_detect - Detect PII entities
pii_anonymize_redact - Remove PII completely
pii_anonymize_replace - Replace with placeholders
pii_anonymize_mask - Mask with asterisks
pii_anonymize_highlight - Highlight PII
pii_models - List available models
pii_health - Check API health

MCP Prompts:

pii-privacy-assistant - Privacy-focused AI assistant
pii-compliance-checker - GDPR/AI Act compliance expert
pii-safe-sharing - Safe data sharing guidance
ai-act-gdpr-lawyer - Legal compliance expert

4. 🔍 MCP Inspector

Features:

Interactive Testing: Test MCP tools and prompts
Debug Interface: Monitor MCP protocol messages
GitHub OAuth: Secure authentication flow
Real-time Logs: Debug server responses

Usage:

npm run inspect

5. 🗄️ Weaviate Integration

Features:

User Tracking: Store GitHub login information
Compliance Data: Track compliance events
Audit Logs: Maintain detailed logs
Vector Search: Semantic search capabilities

Schema:

class GitHubUser {
  username: string
  lastLoginAt: date
}

6. 🎤 Speech-to-Text (OpenAI Whisper)

Features:

Real-time Transcription: Convert audio to text
Multiple Formats: Support for various audio formats
High Accuracy: OpenAI Whisper model
Seamless Integration: Built into input components

Implementation:

const transcribeAudio = async (audioBlob: Blob): Promise<string> => {
  const formData = new FormData();
  formData.append("file", audioBlob, "audio.webm");
  formData.append("model", "whisper-1");
  
  const response = await fetch("https://api.openai.com/v1/audio/transcriptions", {
    method: "POST",
    headers: { Authorization: `Bearer ${apiKey}` },
    body: formData,
  });
  
  return response.json().text;
};

🛡️ Compliance Features

GDPR Compliance

Data Minimization: Collect only necessary data
Purpose Limitation: Use data for stated purposes only
Storage Limitation: Automatic data retention policies
Integrity & Confidentiality: Protect against unauthorized access

EU AI Act Compliance

Risk Assessment: Identify high-risk AI systems
Documentation: Maintain detailed system records
Transparency: Clear AI system information
Human Oversight: Ensure human control over AI systems

Real-time Monitoring

Compliance Score: Overall compliance percentage
Issue Tracking: GDPR and AI Act violations
Trend Analysis: Compliance trends over time
Alert System: Immediate violation notifications

🚀 Deployment

Frontend Deployment (Lovable.dev)

Push changes to your repository
Lovable.dev automatically deploys
Access via your custom domain

Backend Deployment

# Using Docker
docker build -t clyra-api .
docker run -p 8000:8000 -e HF_API_TOKEN=your_token clyra-api

# Using Cloud Run
gcloud run deploy clyra-api --source .

MCP Server Deployment

# Deploy to Cloudflare Workers
npm run deploy

🔧 Development

Local Development

# Start all services
npm run dev:all

# Start individual services
npm run dev:frontend
npm run dev:backend
npm run dev:mcp

Testing

# Frontend tests
cd frontend && npm test

# Backend tests
cd streamlit && python test_fastapi.py

# MCP Inspector
npm run inspect

Environment Variables

See env.example for comprehensive environment configuration.

📚 Documentation

API Documentation: streamlit/README_FASTAPI.md
MCP Inspector: MCP-INSPECTOR.md
Frontend Components: frontend/src/components/
Environment Setup: env.example

🤝 Contributing

Fork the repository
Create a feature branch
Make your changes
Add tests if applicable
Submit a pull request

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🆘 Support

Issues: GitHub Issues
Documentation: See individual component READMEs
API Testing: Use the interactive docs at /docs

🎯 Roadmap

🔒 Clyra - Protecting Privacy in the AI Era

Built with ❤️ using modern web technologies and powered by lovable.dev

Name		Name	Last commit message	Last commit date
Latest commit History 32 Commits
.claude		.claude
frontend		frontend
src		src
streamlit		streamlit
.dev.vars.example		.dev.vars.example
.env.example		.env.example
.gitattributes		.gitattributes
.gitignore		.gitignore
.prettierrc		.prettierrc
ENV.EXAMPLE		ENV.EXAMPLE
LICENSE		LICENSE
MCP-INSPECTOR.md		MCP-INSPECTOR.md
README.md		README.md
image.png		image.png
mcp-inspector.json		mcp-inspector.json
package-lock.json		package-lock.json
package.json		package.json
run-inspector.sh		run-inspector.sh
tsconfig.json		tsconfig.json
worker-configuration.d.ts		worker-configuration.d.ts
wrangler.jsonc		wrangler.jsonc

License

niranjanxprt/Clyra

Folders and files

Latest commit

History

Repository files navigation

🔒 Clyra - AI Privacy & Compliance Platform

🌟 Key Features

🏗️ Architecture Overview

🚀 Quick Start

Prerequisites

1. Clone and Install

2. Environment Setup

3. Start the Services

4. Access the Application

📁 Project Structure

🔧 Component Details

1. 🎨 Frontend (Lovable.dev Power)

2. ⚡ Backend (FastAPI)

3. 🔗 MCP Server (Model Context Protocol)

4. 🔍 MCP Inspector

5. 🗄️ Weaviate Integration

6. 🎤 Speech-to-Text (OpenAI Whisper)

🛡️ Compliance Features

GDPR Compliance

EU AI Act Compliance

Real-time Monitoring

🚀 Deployment

Frontend Deployment (Lovable.dev)

Backend Deployment

MCP Server Deployment

🔧 Development

Local Development

Testing

Environment Variables

📚 Documentation

🤝 Contributing

📄 License

🆘 Support

🎯 Roadmap

About

Resources

License

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Contributors 5

Uh oh!

Languages

Packages