Self Hosted Document Management!

Folks, the first new post since I set this up and it's about a new service.

It's Paperless-ngx. One I'm sure many have heard of over and over on all those threads asking "What's your favorite service no one has heard of" or the like.

Centralized, tag based, AI assisted document management. I fell in love of this idea of tags for information I used to sort and organize into folders when I spun up LinkDing for bookmark management (highly recommend).

I'm slowly migrating a ton of data I have stored in Google Drive over to it - mostly tax documents or insurance. It was a breeze to setup, though I ran into some weirdness where I'd have to redeploy a few times in Docker to get things working - intermittent deploys would just end up with a dead page. I may have to investigate this further if automated updates cause this to happen as well. That's pretty much the only issues I've had though. Otherwise it's been cool (albeit frustrating at the start) to see the application learn what kinds of documents should get each tag. I particularly like that you can configure newly consumed documents to have a tag so you can create a kind of "to be processed" view with just documents that might have some metadata that needs tweaked.

In any case there's a ton of functionality here that I'm sure I'm just starting to uncover so I may make another update post after I have some time to import everything and tune my document library a little bit more.

Anyway as usual I have my configuration information here. Any changes or updates to my deployment configuration will be documented there.

Update 5/21/24 - The following morning I had to redeploy because the website became inaccessible for seemingly no reason (no logs seemed weird anywhere from what I can tell). Feels very similar to what I was seeing yesterday. I will update the README with my config linked above with any information on the fix - I expect the issue is with my setup and not with the service itself.

Update 5/25/24 - The issue seems to have been related to a configuration with my Traefik labels. I decided to just use a file provider for this service and everything has been working swimmingly ever since.