A self-hosted corner of the internet, running local AI, image generation, and enough servers to keep a tuxedo cat warm on cold nights.
This is a personal home lab — a collection of self-hosted AI tools and tinkered-together services running 24/7 on a custom-built PC.
The real power behind the operation is a very serious tuxedo cat who supervises all hardware decisions, judges every failed Docker deployment, and has strong opinions about thermal paste application.
The images on this page were all generated locally — no cloud, no API calls, just a Flux model running on an RTX 5080 and way too much free time.
Open WebUI running on top of Ollama via Docker. Full chat interface with model switching, no data leaving the house.
Ollama + Open WebUIComfyUI with Flux running native CUDA on an RTX 5080. 16GB VRAM means fast, high-quality generations locally.
ComfyUI + FluxTailscale mesh network keeps everything accessible from anywhere. Cloudflare Tunnels for public-facing services with Zero Trust auth.
Tailscale + CloudflareImmich self-hosted photo library. iPhone backups go straight to the home server, not someone else's cloud.
ImmichKokoro TTS running locally for high-quality voice synthesis without sending audio to external APIs.
KokoroAll systems are supervised by one very attentive tuxedo cat. Uptime guaranteed as long as the food bowl is full.
Proprietary