Hey HN,

I wanted to share something I’ve been working on: *RAG-Guard*, a document AI that’s all about privacy. It’s an experiment in combining Retrieval-Augmented Generation (RAG) with AI-powered question answering, but with a twist — your data stays yours.

Here’s the idea: you can upload contracts, research papers, personal notes, or any other documents, and RAG-Guard processes everything locally in your browser. Nothing leaves your device unless you explicitly approve it.

### How It Works: - * Zero-Trust by Design*: Every step happens in your browser until you say otherwise. - * Local Document Processing*: Files are parsed entirely on your device. - * Local Embeddings*: We use [all-MiniLM-L6-v2](https://huggingface.co/sentence-transformers/all-MiniLM-L6-v...) via Transformers.js to generate embeddings right in your browser. - * Secure Storage*: Documents and embeddings are stored in your browser’s encrypted IndexedDB. - * Client-Side Search*: Vector similarity search happens locally, so you can find relevant chunks without sending anything to a server. - * Manual Approval*: Before anything is sent to an AI model, you get to review and approve the exact chunks of text. - * AI Calls*: Only the text you approve is sent to the language model (e.g., Ollama).

No tracking. No analytics. No “training on your data.”

### Why I Built This: I’ve been fascinated by the potential of RAG and AI-powered question answering, but I’ve always been uneasy about the privacy trade-offs. Most tools out there require you to upload sensitive documents to the cloud, where you lose control over what happens to your data.

With RAG-Guard, I wanted to see if it was possible to build something useful without compromising privacy. The goal was to create a tool that respects your data and puts you in control.

### Who It’s For: If you’re someone who works with sensitive documents — contracts, research, personal notes — and you want the power of AI without the risk of unauthorized access or misuse, this might be for you.

### What’s Next: This is still an experiment, and I’d love to hear your thoughts. Is this something you’d use? What features would make it better?

You can check it out here: [https://mrorigo.github.io/rag-guard/]

Looking forward to your feedback!

NSFW Image Generator (free with invite code: HN777)

Nvidia Jetson Thor

How the Richest People in America Avoid Paying Taxes

Ask HN: Why hasn't x84 caught up with Apple M series?

Politically correct language alienates mainstream voters, says Dem think tank

Powerful GPUs or Fast Interconnects: Analyzing Relational Workloads

Show HN: Tool to optimize keyboard layout by counting characters

Show HN: RAG-Guard: Zero-Trust Document AI

Director of Pentagon's Defense Innovation Unit Resigns

An illustrated guide to AWS Security Groups

New AI attack hides data-theft prompts in downscaled images

JavaScript: In what order do these console logs appear

How life-size cows made of butter became an iconic symbol of the Midwest

Are Marathons and Extreme Running Linked to Colon Cancer?

Heavy boron isotopes trace recycled water in deep-mantle reservoir

Show HN: The browser that keeps your projects in order

URL Context

Elon Musk's xAI sues Apple and OpenAI, alleging anticompetitive practices [pdf]

Health insurance prices likely to rise next year

Musk firms sue Apple and OpenAI, alleging they hurt competition

Why L.A. can't back out of hosting 2028 Olympics

EO 14338: Improving Our Nation Through Better Design [pdf]

Intel's Clearwater Forest E-Core Server Chip at Hot Chips 2025

Daily actions can shape how righties vs. lefties process visual input

Improving Search for Deep Research Agents

Biotechs Turn to Digital Coins, Crypto to Boost Stock Prices

Show HN: I built an AI trip planner

It frustrates me that YouTube AI tutorials have become teasers

Farmers Insurance data breach affects 1.1M people

A '90s Goo Goo Dolls Hit Became the Song of the Summer

Ask HN: What problem you are solving as a founder?