Golden Paths in Platform Engineering

A golden path is the supported, recommended way to accomplish a task in your organization. It is not the only way — it is the easy way. The distinction matters.

What a Golden Path Looks Like

When a developer wants to create a new microservice, the golden path might be:

# One command to scaffold a service
platform create service \
  --name order-processor \
  --lang python \
  --template event-driven

# Result: a repo with
# - Dockerfile (multi-stage, security-scanned base)
# - CI pipeline (lint, test, build, deploy)
# - Kubernetes manifests (HPA, PDB, resource limits)
# - Monitoring (dashboard, alerts, SLO)
# - Documentation template

The developer writes application code. Everything else is handled.

Golden Path vs Golden Cage

A golden path offers a paved road. A golden cage forces you onto it.

Golden path: "Here is the standard way to deploy. It handles 90% of use cases. If you need something different, here is how to diverge."

Golden cage: "You must use this template. No exceptions. No customization. File a ticket if it does not work."

Golden cages create shadow infrastructure. Teams that cannot get what they need from the platform build their own tooling, which is worse than having no platform at all.

Designing Effective Golden Paths

Start with the 80% Case

Interview your developers. What takes the most time? Where do they get stuck? The first golden path should address the most common workflow, not the most complex one.

Make It Optional but Attractive

The golden path should be significantly easier than the alternative. If your standard CI pipeline takes 5 minutes to set up and the custom approach takes 2 days, developers will choose the golden path voluntarily.

Provide Escape Hatches

Every golden path needs a documented way to diverge. Teams with legitimate edge cases should not feel trapped. Document the tradeoffs: "If you leave the golden path, you are responsible for your own monitoring and incident response."

Version and Evolve

Golden paths are software products. They need versioning, changelogs, and migration guides. When you update the standard Dockerfile base image, existing services should get the update with minimal effort.

Measure Adoption

Track what percentage of services use the golden path. If adoption drops, the platform is not meeting developer needs. If adoption is forced at 100%, you might have a golden cage.

Common Golden Paths

The Platform Team's Job

The platform team does not tell developers what to build. It makes the right thing the easy thing. When security best practices, observability, and reliability are baked into the golden path, every new service starts with a solid foundation.

That is the power of a well-designed golden path: developers choose quality because it is the path of least resistance.

---

Ready to go deeper? Learn Docker, Ansible, and Terraform with hands-on courses at CopyPasteLearn.