Zero-Trust Document AI

Finally, chat with your documents without sending them to Big Tech. Everything stays in your browser until you hit "send."

Local Processing
Your Data Stays Private
Approve Every Chunk
View on GitHub
100%
Local Processing
0
Data Breaches
MIT
Open Source

How RAG-Guard Works

A transparent, secure approach to document AI that puts you in control

1

Upload Documents

Add your PDFs, docs, or text files. Everything stays in your browser - no upload to our servers.

2

Local Processing

Your documents are chunked and embedded locally using WebAssembly - zero data leaves your device.

3

Review & Approve

See exactly which chunks will be sent with your query. You control what data gets shared.

4

Query Securely

Only approved chunks are sent to the AI provider. Your full documents never leave your browser.

Why Choose RAG-Guard?

Built for privacy-conscious professionals who need AI without compromise

Local AI Processing

Your documents never leave your browser. Everything runs locally using WebAssembly.

  • Client-side document parsing
  • Local embeddings generation
  • No server uploads required

Granular Control

See and approve every piece of data before it's shared with AI providers.

  • Preview all relevant chunks
  • Approve or reject individually
  • Complete transparency

Universal Compatibility

Works with any AI provider - OpenAI, Anthropic, local models, and more.

  • OpenAI integration
  • Anthropic Claude support
  • Local model compatibility

Modern Web Experience

Built with cutting-edge web technologies for a smooth, native-like experience.

  • Progressive Web App
  • Responsive design
  • Offline capabilities

No Subscription Lock-In

Own your AI workflow forever. Self-host for free or pay only for the AI models you use.

  • Free to self-host
  • MIT open source license
  • Pay-per-use AI costs only

Enterprise Ready

Meets compliance requirements with audit trails and air-gapped deployment options.

  • HIPAA/GDPR compliant
  • Full audit logging
  • Air-gapped deployment

Perfect For Privacy-Conscious Professionals

When your data can't afford to be compromised

Healthcare Professionals

Analyze patient documents and research papers while maintaining HIPAA compliance. Your sensitive data never leaves your device.

Legal Teams

Review contracts and case documents with AI assistance while maintaining attorney-client privilege and confidentiality.

Financial Analysts

Process financial reports and market research without exposing sensitive client information to external AI services.

🔒 Zero-Trust Architecture

Your documents are processed entirely in your browser. We can't see your data even if we wanted to - it never reaches our servers.

Quick Start

Get RAG-Guard running in minutes with Docker or set up a development environment

git clone https://github.com/mrorigo/rag-guard.git
cd rag-guard
docker-compose up -d
open http://localhost:3000
git clone https://github.com/mrorigo/rag-guard.git
cd rag-guard/backend && npm install && npm run dev
cd ../frontend && npm install && npm run dev
View On GitHub