https://teerakarna.github.io/tags/operators/ Recent content in Operators on Bertie's Blog Hugo en-gb Sun, 13 Sep 2026 16:25:01 +0700 https://teerakarna.github.io/posts/candor-3-putting-an-llm-on-a-leash/ Mon, 28 Sep 2026 10:10:00 +0700 https://teerakarna.github.io/posts/candor-3-putting-an-llm-on-a-leash/ Adding a model to a Kubernetes operator is the easy part. The work is the interface, the gate, the budget ceiling, and making sure untrusted scanner output cannot talk to the model as if it were you. https://teerakarna.github.io/posts/candor-2-operator-pattern-under-constraint/ Sun, 20 Sep 2026 19:55:00 +0700 https://teerakarna.github.io/posts/candor-2-operator-pattern-under-constraint/ CRDs, the provider pattern, and a content-addressed fingerprint that decides whether an expensive model call happens at all. Learning the operator pattern by building one where getting it wrong costs money. https://teerakarna.github.io/posts/candor-1-decisions-before-the-code/ Sun, 13 Sep 2026 16:01:00 +0700 https://teerakarna.github.io/posts/candor-1-decisions-before-the-code/ Before writing a single controller, I spent the time on research, positioning, naming, and governance. Here is what that bought, and why it matters more for an AI tool than a normal one.