Self-Hosted AI & Autonomous Agent Systems Engineer
I build production AI that runs on your own hardware — not someone else's cloud.
In Production
Real businesses running on the platform right now — a storefront out front, a self-hosted LMS behind it.
InkyPyrus
The AI content studio that runs on the platform.
AfrEcoSoil
Biological soil-rehabilitation knowledge, sold and taught online.
PtyLaunchpad
South African company-formation & compliance courses.
Hermes Agent Forge
A hands-on training program for the Hermes Agent stack.
I run a 3-node self-hosted AI platform — local LLM inference with web-search augmentation, a GPU media/TTS node, and a containerized multi-site hosting estate behind Cloudflare — orchestrating autonomous agents that build and run real businesses.
- Self-hosted 27B LLM via llama.cpp
- Web-search-augmented proxy + model hot-swap
- Hermes gateway (Telegram / WhatsApp)
- Remotion render · agent control plane
- 5 containerized WordPress stacks
- Caddy reverse proxy
- Cloudflare tunnels · zero open ports
- Neural TTS pipeline
- Twilio + x.ai voice call-screener
- Moodle LMS + test harness
What I build
Featured builds
A content studio running on the platform.
The Self-Hosted Course BusinessSelling knowledge online usually means renting three businesses from three vendors: a storefront, a learning platform, and — increasingly — an AI that reads your own course material back to you off someone else’s server, metered by the token. I went the other way on the two that matter. The shopfront out front can be […]
The Call ScreenerTwilio + x.ai voice agent that screens calls and sends a Telegram transcript.
The Hosting Estate5 containerized WordPress sites behind Caddy + Cloudflare.
The Self-Hosted AI PlatformA 3-node self-hosted AI platform — local LLM inference, a GPU media/voice node, and a containerized hosting estate — that builds and runs real businesses at near-zero inference cost.
HermesTelegram/WhatsApp gateway into local models.