Can Kubernetes Run AI Inference in Your Homelab? An Honest Answer, with Three CNCF Projects
The homelab inference talk: three mismatched GPUs, one Kubernetes cluster, and an honest answer about whether HAMi, Kaito and llm-d beat running Ollama.
Speaking
I speak about Kubernetes, virtualization and AI inference, mostly from the position of someone who runs the stack at home before recommending it at work. Most talks end up here in more detail than fits in thirty minutes.
Talks
The Future of AI Inference is Small, Distributed, and Budget-Conscious
A lightning talk arguing that the next decade of AI infrastructure is small models at small organisations, not frontier clusters.
Where I have spoken
- KCD UK Edinburgh 2026
- Cloud Native Summit 2026
- KCD Czech & Slovak 2026