"Platform engineering" went from a conference concept to a standard job title in about 18 months. Before you add it to your org chart, it is worth having a clear definition of what problem you are actually solving.
DevOps was a culture change, not a job title: developers should be responsible for their own deployments, monitoring, and on-call rotations. In practice, DevOps became a job title meaning someone who manages CI/CD pipelines, cloud infrastructure, and deployment tooling.
Platform engineering is a team that builds and maintains the internal tooling that other engineering teams use. The goal is reducing cognitive load on product engineers by abstracting away infrastructure complexity. The Internal Developer Platform (IDP) is the artifact this team produces.
Fewer than 20 engineers. Deployment and infrastructure do not consume significant engineering time. Infrastructure is standard: managed databases, a single cloud provider, a straightforward CI/CD setup.
20 to 30+ engineers with teams solving the same infrastructure problems independently. Setting up a new service takes more than two hours of infrastructure work before a line of product code is written. You are starting to see toil — repetitive operational work that scales with team size.
Before building custom, evaluate Backstage (open source), Cortex or OpsLevel (commercial), or simply a more opinionated standard toolchain. Many companies solve the coordination problem with convention and documentation rather than custom tooling.
Axented helps growing engineering teams structure their infrastructure and DevOps practices. → axented.com/platform-infrastructure
%20(1)-2.svg)
