Build a complete Paperless-ngx stack in Docker and take control of your documents. We’ll get Paperless running first (works great on its own), then optionally add local AI with Ollama + Open WebUI and upgrade OCR using Paperless-GPT and Paperless-AI for more accurate, searchable text - no cloud required.

Full Guide Here: https://technotim.com/posts/paperless-ngx-local-ai/
Code Here: https://github.com/timothystewart6/paperless-stack

Paperless NGX: https://github.com/paperless-ngx/paperless-ngx
Paperless AI: https://github.com/clusterzx/paperless-ai
Paperless GPT: https://github.com/icereed/paperless-gpt

Merch Shop 🛍️: https://l.technotim.com/shop
Support me on Patreon: https://www.patreon.com/technotim
Sponsor me on GitHub: https://github.com/sponsors/timothystewart6
Subscribe on Twitch: https://www.twitch.tv/technotim
Become a YouTube member: https://www.youtube.com/channel/UCOk-gHyjcWZNj3Br4oxwh0A/join
Gear Recommendations: https://l.technotim.com/gear
Get Help in Our Discord Community: https://l.technotim.com/discord
2nd channel: https://www.youtube.com/@TechnoTimTinkers

(Affiliate links may be included in this description. I may receive a small commission at no cost to you.)

00:00 - Intro - Self-hosted, private, AI optional
00:30 - What's included in this Paperless-NGX Stack?
01:11 - Paperless Architecture (Paperless ↔ AI ↔ Ollama)
01:46 - Stack Overview - All services explained 
02:45 - Folder layout + ports + .env files
03:10 - Docker Compose up
03:23 - Ollama + Open WebUI setup (port 3001)
03:59 - NVIDIA GPU quick test
04:22 - First login: Paperless-ngx setup (port 8000)
04:43 - Create Paperless API token
05:21 - Paperless-AI setup + settings (port 3000)
08:17 - Upload docs + baseline OCR and metadata results
11:30 - Processing documents with Paperless-AI
12:50 - Updated metadata from Paperless-AI
14:39 - RAG Chat with your documents
16:09 - OCR is still not good
16:31 - Paperless GPT replaces OCR with a Vision model
16:51 - Paperless-GPT setup (port 3002)
17:25 - minicpm-v - an open source multimodal vision model
18:05 - Download opencpm-v to Ollama
18:22 - Paperless-GPT walkthrough (port 3002)
19:10 - Vision OCR demo (before/after)
21:05 - Vision OCR with Watermarks (before/after)
22:32 - Automate with Paperless workflows (auto tags)
24:40 - Testing automated workflow with Vision OCR
26:18 - My thoughts about PaperlessNGX with Local AI

Thank you for watching!

YouTube

Paperless-ngx is a self-hosted document management system that automatically performs OCR on scanned PDFs, organizes them with tags and correspondents, and makes everything searchable. This tutorial demonstrates a complete Docker-based setup, including optional local AI integration using Ollama for enhanced OCR accuracy through vision models (minicpm-v), automated metadata tagging with paperless-ai, and RAG-based document chat capabilities. The setup keeps all data local without cloud dependencies, with the AI components being entirely optional for those who only need basic document management and OCR.

Paperless-ngx + Local AI (Optional): Better OCR, Self-Hosted, No Cloud