The CI/CD Roadmap

Why manual releases fall apart, the pipeline model every CI platform shares, and the build systems that actually turn source code into something deployable.

What actually starts a pipeline, what a real build-and-test stage looks like, and the two techniques that make slow pipelines fast again.

Workflows, secrets, reusable pieces, and your own runners — going from Topic 0.4's minimal example to genuinely production-grade GitHub Actions usage.

The same pipeline model from Phase 0.2, now expressed as a controller dispatching work to agents, written in a Jenkinsfile committed to your own repository.

One YAML file, deeply integrated with GitLab itself — the third major platform's own take on the exact same pipeline model.

The build produces something real — a JAR, a container image — and an artifact repository is where it lives, gets versioned, and moves toward production deliberately.

Every deployment strategy, expressed as actual pipeline stages — and the organizational maturity question that actually decides how far to automate.

A pipeline that can deploy your application can, if compromised, do just as much damage as a compromised application itself — this phase covers protecting it accordingly.

A test strategy pipelines can trust (and what to do about flaky tests), CI for monorepos that only builds what changed, automated versioning and release notes, and measuring delivery with DORA metrics and pipeline observability.