Where the reliability habit came from.

An experience snapshot of employed platform work, drawn as three horizontal bands with a narrow column down the right side. The first band, labelled runs on, shows hybrid infrastructure as two boxes side by side: Cloud, described as capacity that moves with demand, and On premises, described as capacity that has to be planned. A two headed connector joins them, marked hybrid, because the estate spans Amazon Web Services and on-premises hardware at the same time. The second band, labelled gets there by, shows continuous integration, continuous delivery and release engineering as a pipeline of four numbered steps reading left to right: build, then test, then release, then deploy, each joined by an arrow. The third band, labelled is watched by, shows observability feeding an incident path: four small boxes in a row, metrics, logs, alert and on-call, joined by light connectors, and then one bold arrow onward into a highlighted box called root cause analysis. From root cause analysis a long arrow travels upward and back to the left, re-entering the first step of the pipeline, carrying the note that the fix goes back through the pipeline, so the loop closes. The column on the right, outlined in the accent colour, is headed disaster recovery and holds two short lines: a way back, designed at the start, and rehearsed rather than documented. A line along the bottom reads: around the clock production, an outage here is public, and there is no quiet window.