CI/CD Pipelines

What and Why

You deploy code manually. That means: SSH into a server, pull the latest, restart the process, cross your fingers. It works until it doesn't — a bad deploy at 5pm on a Friday, no rollback plan, logs that tell you nothing.

CI/CD (Continuous Integration / Continuous Delivery) pipelines automate the path from code commit to running production software. CI validates every change automatically. CD gets validated changes into production — either automatically (Continuous Deployment) or with a manual gate (Continuous Delivery).

In production systems this matters because:

• Manual deploys don't scale across teams and services
• Human steps introduce inconsistency — what ran on the developer's machine ≠ what runs in production
• Fast feedback loops catch regressions before they compound
• Audit trails, reproducibility, and rollback become structural guarantees instead of heroics

You encounter pipelines when onboarding to any team with more than one engineer, when debugging why a deploy failed silently, or when designing the path from PR to production for a new service.