ViewTube

Everyone's got an opinion about AI in ops. Most of them have never taken a 3 AM page. I've spent years as a senior platform/SRE engineer: real on-call, real outages, real production Kubernetes. Now I'm working through the question everyone's dancing around: how does AI actually fit into DevOps and platform engineering without breaking prod at 3 AM? What you'll find here: LLM agents triaging real incidents and the exact moments they confidently get it wrong. Human-in-the-loop workflows a serious engineer would actually let near production. The security, the approval gates, the platform glue that makes it real. No vendor demos, no edited-out failures. If it breaks, you watch it break. For platform, SRE, and DevOps engineers who already know their way around a cluster and asking the sharper question: am I going to be the one wielding AI in ops, or the one automated out of the room? If that's you, you're in the right place. New videos regularly. Subscribe and let's build this properly.