Join me as I share what I learn building Azure infrastructures, automation workflows and AI integrations — from real projects, not slides.
01 — Use cases
Write @claude in a GitHub issue and get a pull request with a plan, a design check and screenshots; merging publishes the site. About an hour to set up: a Claude Console workspace with a spend cap, a federation rule instead of an API key, four repo variables and one workflow.
Every Bicep module in your repo gets one version that CI checks on every pull request, modules come from a private registry, and each PR passes a version check, PSRule and what-if before review. The version gate takes about an hour to set up; all scripts included.
A 90-day plan to roll out AI tools in an IT organisation: one-line mission, data traffic light, baseline, champions with protected time, staged training and four numbers that show whether it pays. With what Shopify, Siemens, UBS and others did.
02 — Latest posts
Every agent PR carries a plan and waits for your approval, using only GitHub's own settings: a required plan, a read-only token, a human approval gate, allowlisted MCP tools and hooks that say no.
Know whether the EU AI Act applies to your Swiss company and what to do first, with a checklist. Where the Act stands in September 2026, and how Switzerland regulates AI without an AI act. My research notes, not legal advice.
Push a Bicep file to GitHub and about 30 seconds later it is a LinkedIn post, with a caption and the code as an image. n8n writes the caption, renders the code and posts both. An experiment, not a production tool.
