Media Stack

The Media Stack started as a joke. I’d become the unofficial movie guy for friends and family, and I got tired of manually hunting down, downloading, and organizing whatever they asked for. So I automated the whole thing.

Now people request what they want through a portal, and the system handles the rest: it searches for the title, grabs it through a download client, the *arr apps sort and rename it into the library, and Plex refreshes so it’s ready to watch. No step in the middle needs me.

Why it mattered

On the surface it’s a media server. Under the hood it was one of my earliest real automation projects, and it taught the fundamentals I lean on everywhere now: connecting services into one pipeline, managing the dependencies between them, troubleshooting failures across systems, and building something that runs without a babysitter. A lot of what I later applied to AI automation started here.

Reliability

Once it’s configured, it mostly runs itself. Requests come in, downloads happen, the library updates, and people watch, all in the background. It breaks once in a while, and when it does the culprit is almost always connectivity or the VPN, not the apps. That’s the real lesson of automation: building it is the easy part, building it so it keeps working when you stop paying attention is the job.