Two weeks ago, Model Context Protocol shipped its 2026-07-28 specification revision.

The biggest updates in this release are: no more stateful initialize handshake, Mcp-Session-Id headers have been retired, and every request …

I created a skill to save users 40-60% on token costs when migrating AI inference workloads from a serverless platform (like Cloud Run or Gemini Enterprise Agent Platform) to GKE. So what is it? What isn’t it? Why a …

Introduction to Distributed RL Sandboxing on GKE

Reinforcement Learning (RL) is the cornerstone of modern AI training. Rather than train a model to produce an expected output, we verify if it has achieved a particular outcome. This is …

Ok, so you’re writing an AI agent to automate some task (who isn’t these days?). But where does it live? That largely depends on what you want it to do.

XPK Cluster Create

Simplifying AI Infra Management

You’ve probably used terraform and kubectl to set up infrastructure and deploy code to it (if you haven’t, thanks for stopping by, but this post might not be super interesting …

I have been hooked on Factorio for years now. It is my favorite video game, and probably one of the most impressively documented feats of software engineering in gaming. Check out their long-lived series of technical blog posts: Factorio …

Backlogged Pinball

Earlier this year, I built a pinball-based demo for Cloud Next 2024. I’m preparing some posts about specific aspects of it, but I wanted to post a brief overview first. Backlogged is a physical pinball machine …