What Production Really Means

At Cerner, my code was used by doctors in hospitals. That changes how you think about “production.”

The shift

In most software, a bug means a frustrated user and a Slack alert. In healthcare software, a bug could mean a doctor does not see a critical lab result. The stakes are qualitatively different.

This taught me something that I carry to every project since: the definition of “production-ready” is set by the people who depend on the system, not by the people who build it.

What changes when stakes are high

1. You test differently. Not just “does it work?” but “does it fail safely?” Happy path testing is table stakes. Edge case and failure mode testing is the actual work.

2. You deploy differently. Feature flags, gradual rollouts, instant rollback capability. The ability to undo a deployment in seconds is not a nice-to-have.

3. You monitor differently. If you cannot see it, you do not understand it. Dashboards are not vanity metrics — they are your only window into whether the system is doing what you think it is doing.

4. You communicate differently. Incident response is a skill. Blameless postmortems are not just a process — they are how you build a team that gets better instead of one that gets defensive.

The transferable lesson

Even when the stakes are lower — a SaaS app, an internal tool, a portfolio site — the mindset transfers. Build like someone depends on it, because they do.