Every so often, the thing you build ends up changing how you work. This may also echo the paradigm of "the obstacle becomes the way," maybe, I don't know that answer yet, still this was useful to m...
Summary: The previous post ended with a dashboard that could tell me what was running. This one is about not having to look at it. Six light panels on the wall, one per machine, turning system stat...
Summary: Some services in this lab run as plain systemd units. Others are moving to a small Kubernetes cluster. That split is permanent and deliberate, and I want to explain why - because the natur...
Summary: The main server in this lab stores everything on a striped array with zero redundancy. If either drive fails, all of it is gone at once - no rebuild, no degraded mode, nothing to recover f...
Summary: There's a moment in every growing homelab where you lose track. Not of anything dramatic - you just cannot say, off the top of your head, whether all the things you built are currently run...
Summary: The lab now has three machines that can run models, and I want to be precise about how that happened, because "three inference hosts" sounds like a plan and it was nothing of the kind. Eac...
Summary: This is the first post in a series about Resonance Lab - a home AI lab that grew, without much of a plan, from one small machine into four, three network segments, and a couple of dozen se...
Summary: In my home AI lab, I’m adding AnythingLLM not because I needed another chatbot, but because I wanted to dogfood the same architecture patterns I evaluate professionally: context boundaries...
Summary: So I'm doing the AI path, like everyone else, and after spending months building out my home lab with dedicated hardware, Ollama and local LLMs, I kept noticing something: the same Qwen mo...
Every so often, a learning journey turns into something a little more project-shaped than you originally planned. "Resonance Lab," my home brew project moniker for the AI lab I built, started that ...