From DevOps tools to internal developer platforms: how platform engineering works
Every team picking its own DevOps tools created sprawl. Platform engineering puts the common pieces behind one self-service platform.

DevOps gave development teams responsibility for building, deploying and running their own software. Many companies took that to mean each team should choose and maintain its own tools. The result was dozens of slightly different pipelines, monitoring setups and cloud configurations, and developers spending a large share of their time on infrastructure rather than features.
Platform engineering is the response. A platform team builds an internal developer platform: a set of shared, self-service tools and templates that product teams use to create, deploy and run software the approved way.
What an internal developer platform includes
| Layer | What the platform provides |
|---|---|
| Developer portal | A catalog of services, owners and documentation, plus templates to start new services |
| Templates ("golden paths") | Ready-made project setups with CI/CD, security scanning and monitoring built in |
| CI/CD | Standard pipelines teams can reuse |
| Infrastructure | Self-service environments and databases with guardrails |
| Observability | Logging, metrics and tracing set up by default |
| Security and compliance | Policies applied automatically, not by review |
Backstage, an open source developer portal created at Spotify and now a Cloud Native Computing Foundation project, is the most common starting point for the portal layer. Commercial portals and cloud provider tools cover similar ground.
Treat the platform as a product
Platforms fail when they are mandated rather than wanted. Successful platform teams treat developers as customers: they find the most painful tasks, make those easy first, measure adoption and keep the golden path optional but clearly better.
The evidence
DORA's 2025 report on AI-assisted software development found that AI's primary role is as an amplifier of an organization's existing strengths and weaknesses, and that the biggest gains come from the system around the tools. A good internal platform is part of that system: it gives AI-generated code a tested, secure path to production instead of a manual one.
When it makes sense
A company with two or three development teams doesn't need a platform team, but it benefits from shared templates and one standard pipeline. Past roughly ten teams, the duplication cost usually justifies dedicated platform staff.
What IT teams should do
- Count how many different CI/CD, monitoring and deployment setups your teams run.
- Ask developers which infrastructure tasks take the most time, and fix those first.
- Build one golden-path template for new services, with security and monitoring included.
- Keep a service catalog showing every service and its owner.