Field notes on Kubernetes optimization, GPU economics, and what we learn beneath production clusters. We publish sparingly, and only when there's something worth saying.