State of my lab before I pack it up for college

Hi everyone!
I’ve been into this hobby for just over a year now. I’m planning to take my rack to college for my last year of computer science undergrad. I’ve grown to depend on stuff like Home Assistant, and the utility of the Gl.inet router for stuff like AirPlay, having your own printer, etc while not missing out on the campus network speeds. I’m also interested in setting up a receipt processing/expense splitting solution for my housemates and I, leveraging the GPU in my rack for local AI vision/OCR — if you have any suggestions for this please LMK, I’m looking into Receipt Wrangler.

Overview of Gear, from top to bottom:

GL.inet Slate 7, with 120mm fan mounted underneath

Netgear GS308EP, JetKVM PoE edition, WD Elements Desktop 8TB (with custom 3D printed mount)

Firebat MK2 Plus Mini PC (Intel N5105, 16GB RAM, 512gb SSD), SONOFF Zigbee Dongle, ScreenBeam MoCA 2.0 adapter

2 x 12TB Toshiba N300 HDDs in RAID1 array

AsRock B460M ITX motherboard w/ Intel i9-10850k, 32gb RAM, 1TB SSD, RX 6600XT 8GB, with custom 3D printed front cover

There’s also a Raspberry Pi 5 4GB wedged in between the 80mm front fan and GPU. It runs qdevice, NUT, and PBS.

1U panel with EVGA 650W Gold SFX PSU & various other power bricks/misc

On the back, there’s just cable management prints, 2x more 80mm fans, and a 120mm fan to ensure everything stays relatively cool.

UPS is a APC Back UPS 1500 VA Pro

Etsy Dark Souls bonfire lamp w/ Zigbee IR blaster for ambiance

What I’m running:

Media: Jellyfin+arr stack, Audiobookshelf/LazyLibrarian, Komga/Suwayomi,

Photos/Files: Immich, Obsidian Selfhosted Livesync, Syncthing, Seafile

Networking: AdGuard Home, Unbound DNS, Caddy reverse proxy, dedicated pia-tun LXC for exit node to dodge some more ads on mobile

AI: Hermes Agent, llama-swap, MemPalace connected to my Obsidian Vault, Openwebui

And Tailscale on everything for remote access.

That’s all, thanks for reading if you got this far!