About Me & This Lab

# Philosophy

I believe the best way to learn infrastructure is by running it. Not in a sandbox, not on a free-tier cloud account that vanishes when you graduate — at home, where downtime costs patience instead of money and every broken service teaches something real.

My approach: containerize everything, automate what repeats, document what breaks. I prefer open-source tooling wherever possible, self-hosted over SaaS, and local inference over API calls when the latency is tolerable. The lab exists to experiment without asking permission from a billing dashboard.

# Origin Story

It started with a Raspberry Pi running Pi-hole in 2019. That single board turned into three, then a mini-PC I bought on eBay for $80 that became my first always-on server. Docker arrived next, followed by Portainer, and suddenly I had something resembling infrastructure instead of just scripts.

The real turning point was moving to local AI inference — running models locally meant building a proper GPU rig at home, which in turn required reliable networking, storage that doesn't fill up with checkpoints overnight, and monitoring so I'd know when something melted (literally). That's what pushed this from hobbyist tinkering into actual homelab territory.

# The New Home

Moving to a new house in 2024 meant starting over — no more borrowed closet space, but also no pre-existing rack. I got to design the network from scratch: proper wiring, UniFi access points throughout, and a dedicated room for servers that actually has HVAC instead of just hoping the hallway stays cool enough.

The goal is a resilient two-node setup with shared storage, automated backups, and everything behind Cloudflare Tunnels so nothing touches port forwarding on the router. This site tracks progress toward that ideal — some days it's close, other days I'm SSH'd in at 2am because Docker decided to restart every container simultaneously.