Who We Are
Self-Hosted Tech is built by people who actually run AI infrastructure at home. Not a rented VPS with Ollama installed once for a blog post — a real homelab with multiple VMs, GPUs, local models, voice synthesis, image generation, and automation that runs 24/7.
What We Do
We teach you how to run AI on your own hardware. No cloud fees. No data leaving your machine. No rate limits.
- Local LLM guides — Ollama, LM Studio, llama.cpp, model selection
- Hardware guides — GPU buying guides, build recommendations, VRAM optimization
- AI tools — ComfyUI for images, Kokoro for voice, n8n for automation
- Homelab tutorials — Proxmox, Docker, networking, storage, security
- Self-hosted alternatives — Replace Google, ChatGPT, and paid SaaS with your own infrastructure
Our Infrastructure
- Proxmox virtualization host
- 7 VMs (gateway, agents, local LLM inference, media, automation, storage, monitoring)
- Multiple GPUs for LLM inference and image generation
- Local TTS for voice synthesis
- n8n for AI workflow automation
- TrueNAS for storage and backups
- All behind a firewall. Nothing exposed to the public internet.
The Philosophy
Self-host your AI. Don’t expose your homelab.
We use Cloudflare in front of everything public. The good stuff stays behind the firewall. This site is statically generated and served from a CDN — not hosted on our infrastructure. That’s intentional.
Connect
- YouTube: Self-Hosted Tech
- GitHub: github.com/selfhostedtech
- RSS: Subscribe