Chapter 3. Orchestrate, Don’t Just Automate: How Platforms Really Work
As platform engineering gains momentum in an organization, many teams begin by combining familiar tools: a developer portal like Backstage and a workflow engine like Argo Workflows or a CI/CD pipeline such as GitHub Actions. These tools are powerful and well understood, and they create the appearance of a mature platform. But appearances can be deceiving.
This setup often leads to what we call the platform façade: a thin layer of UI and automation that looks like a platform but lacks the underlying coordination, governance, and lifecycle management to support real-world scale. It enables short-term wins, but long-term complexity and fragility build up beneath the surface.
Escaping the Platform Façade with Orchestration
Without orchestration, such platforms decay. Let’s now explore what exactly we mean by “platform orchestration.”
The Limits of Pipelines and Workflow Engines
CI/CD tools and workflow engines automate step-by-step tasks. They are excellent for executing deployments, handling retries, or coordinating parallel actions. But they are fundamentally transient systems. They do not manage persistent state, enforce policies over time, or track ownership.
For example, a pipeline can deploy a database or an application, but it cannot tell you who owns it, whether it complies with current policies, or whether it is still in use six months later. When teams rely solely on pipelines for provisioning, they ...
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Read now
Unlock full access