DevOps Team Structures in 2025: Platform Engineering, Embedded SREs, and the Role of Automation

The DevOps movement started with "you build it, you run it" — the idea that development teams should own their services end-to-end, including operations. In practice, most organisations found that pure "you build it, you run it" creates inconsistency: every team invents its own deployment process, its own monitoring setup, its own infrastructure patterns.

The pendulum has swung back toward specialisation, but in a more nuanced form. Platform engineering — building and maintaining the internal tooling that application teams use to ship — has emerged as the dominant model for organisations above a certain scale.

The Platform Engineering Model

A platform engineering team (sometimes called an Internal Developer Platform team or Enablement team) builds and maintains the "paved road" — the opinionated, supported path that application teams follow to provision infrastructure, set up CI/CD, manage deployments, and get monitoring.

The platform team doesn't run application teams' workloads for them. They provide the tools, templates, and processes that make it easy for application teams to do it themselves correctly. The platform team owns the platform; the application teams own their products.

Where QuickInfra Fits

QuickInfra is infrastructure for a platform team to offer to application teams. The platform team configures QuickInfra with:

• Organisation-level security policies and compliance requirements
• Approved infrastructure templates for standard application architectures
• Pre-configured CI/CD pipeline templates for each supported stack
• Connected cloud accounts with appropriate permission scoping per team

Application teams get a self-service interface to provision environments, set up pipelines, and manage deployments within the guardrails the platform team established. The platform team doesn't become a bottleneck for every infrastructure request.

Embedded SREs

Some organisations embed SRE resources within application teams rather than centralising them. This model maximises context — the SRE deeply understands the specific application — but creates consistency challenges across teams.

QuickInfra supports this model by standardising the infrastructure layer so embedded SREs don't each reinvent the same patterns. Two embedded SREs in different teams using QuickInfra will end up with consistent infrastructure configurations because they're working from the same templates and policies.

The Right Model for Your Scale

Startups and small companies: one person handling all infrastructure, QuickInfra handling the automation. Mid-stage companies: a small platform team (2-4 engineers) using QuickInfra as their primary tooling. Large organisations: a full platform engineering team building on top of QuickInfra alongside other internal tooling.