After this lesson, you will be able to: Explain what CI/CD is, why it exists, and the developer-experience problem it solves.
CI/CD turns 'push to main' into 'live in production' automatically. The pipeline is the most-leveraged piece of infrastructure your team owns.
This is a free introductory lesson. No purchase required.
CI, Continuous Integration: every commit triggers automated builds + tests. Catches breakages within minutes instead of weeks. CD, Continuous Delivery: every passing build is AUTOMATICALLY deployable, but actual deploy may need a human click. CD, Continuous Deployment: every passing build is automatically deployed to production. Higher bar; requires strong tests + observability. Most teams say 'CI/CD' meaning 'CI + Continuous Delivery'. True Continuous Deployment is the senior shop's flex.
Manual deploys: a person SSHs in, runs commands, hopes. Takes 30+ minutes; can't deploy at 11pm; mistakes are common. No CI: bugs reach `main` and you find them during deploy or worse, in production. Slow feedback loops: 'works on my machine' lasts weeks. Burnout: humans become the bottleneck and the single point of failure.
On every PR: lint + type-check + test + build (don't break main). On merge to main: build a versioned artifact (Docker image, npm package, jar). Tag with the commit SHA. Push to registry. Deploy artifact to staging automatically. Run smoke tests. On release tag (or manual approval): deploy the same artifact to production. If a deploy fails: automatic rollback or alert the on-call.
GitHub Actions: native to GitHub, the default for new projects. We use it in this sub-track. GitLab CI: built into GitLab. Same shape; YAML in `.gitlab-ci.yml`. CircleCI / BuildKite / Drone: dedicated CI platforms; often faster for big repos. Argo CD / FluxCD: GitOps deploys for K8s. Jenkins: the elder statesman; still everywhere in enterprises. New projects rarely pick it.
Pick the senior answer.