Two and a half years ago, I started a new job. I was very excited, and I looked forward to making my mark in the first 90 days before tackling any big issues. A few weeks into this new job, the system administrator began complaining about having to explain yet again the steps to deploy a release. I started asking questions. What I discovered shocked and appalled me, and it spurred me to dream a dream. This sys admin described a process for deploying a release to production, for even tiny code changes, that involved three repetitions of manual regression tests, putting the entire site into maintenance mode, working between midnight and 4:00 a.m., and other nightmarish procedures. I embarked on a quest to bring this organization into the 21st century of infrastructure and development best practices. In broad strokes, this involved organizational changes in the department by creating new roles and reporting relationships, motivating key individuals, introducing new development practices, and, of course, introducing new technologies. Our experiences show three key results: motivational changes are tightly aligned with organizational changes, ops and dev working collaboratively is mandatory, and technology choice matters least of all.