Hey Product Hunt! 👋

I got tired of "chat with your PDF" tools just being thin cloud API wrappers. Uploading private client details or NDAs to a third-party server shouldn't be the only option.

I built DocMind to solve this. It is a 100% offline document chat mobile app powered by a local, on-device RAG engine.

What it does:

Absolute Privacy: Your files never leave your phone. Zero cloud APIs, zero data leaks.
Advanced Local RAG & OCR: Combines neural vector embeddings with lexical search, and actively reads scanned images locally.
Anti-SaaS Pricing: Fully functional free tier (3 uploads, 8 messages/chat). For unlimited access, bypass subscription fatigue and own it forever for a one-time $19.99 (or $1/mo).

For the technical folks: Running a local LLM on mobile without melting the battery is a massive challenge. I managed the SoC thermal envelope using Vulkan GPU compute, strict CPU thread-pinning to mid-performance cores, and micro-sleeps during indexing.

I'd love your feedback on the inference speed and UI. I'll be in the comments all day to answer questions about local mobile AI!

Check it out here: https://docmind.si