You can autoscale infrastructure. You can't autoscale judgment.

Most engineering processes aren't processes — they're habits. If only one person can do it, it's a dependency. If you can't test it, you can't improve it. I took DevOps principles — repeatability, testability, transparency — and applied them to the human side of scaling: hiring, onboarding, incident response, architecture decisions.

At Axios, we rebuilt the CMS during an election cycle. At Al Jazeera, we scaled video infrastructure under breaking-news pressure. At The Atlantic, we modernized legacy systems while keeping the lights on. None of that succeeded because of tools. It succeeded because we made decisions visible, failures learnable, and knowledge transferable.

This talk walks through real architecture decisions, postmortems, and hiring processes from those three newsrooms. I'll show you the frameworks that worked when mistakes were public and expensive. Live examples on screen — ADRs, incident timelines, interview rubrics.

You'll leave able to identify single points of failure in people (not just systems), test your own processes the way you test code, and build transparency into decision-making so teams scale judgment, not just infrastructure.