Building an AI application is only the first step. The real challenge begins after the model is trained and integrated into a system. Code changes, model updates, data pipeline modifications, and infrastructure changes must all be deployed safely and reliably. Without proper automation, releasing updates quickly becomes slow, error-prone, and difficult to manage.

Continuous Integration and Continuous Deployment (CI/CD) bring discipline and automation to this process. CI ensures that every change to your codebase is automatically tested and validated. CD extends this pipeline to automatically package, deploy, and release updates to staging or production environments.

For AI applications, CI/CD pipelines often go beyond traditional software workflows. They may include steps such as validating data pipelines, running model evaluation tests, versioning models, building containers, and deploying inference services.

In this chapter, you will learn how CI/CD pipelines are designed for AI systems.