Luca BertonPlatform Engineering in 60 Seconds
Platform engineering is the discipline of building and maintaining internal developer platforms (IDPs) — self-service tools and infrastructure that enable development teams to build, deploy, and operate software without needing to understand the underlying infrastructure complexity.
Think of it as building a paved road for your developers. Instead of each team figuring out how to deploy to Kubernetes, configure networking, set up monitoring, and comply with security policies, the platform team provides all of that as a self-service product.
Platform Engineering vs DevOps
| Aspect | DevOps | Platform Engineering |
|---|---|---|
| Philosophy | "You build it, you run it" | "We build the platform, you build on it" |
| Developer burden | High (own entire stack) | Low (platform handles infrastructure) |
| Standardisation | Varies per team | Golden paths with opt-out |
| Scaling | Doesn't scale well past ~10 teams | Designed to scale (platform team:dev teams = 1:10+) |
| Compliance | Enforced per team (inconsistent) | Built into the platform (consistent by default) |
Key insight: Platform engineering isn't anti-DevOps — it's the evolution of DevOps at scale. When "you build it, you run it" creates 50 teams each reinventing deployment pipelines, something needs to change.
Core Concepts
- Internal Developer Platform (IDP): The product that platform teams build. Includes self-service infrastructure provisioning, CI/CD, observability, security scanning, and documentation — all integrated.
- Golden Paths: Pre-paved, opinionated workflows for common tasks. "Here's how you deploy a service" with best practices baked in. Developers can deviate, but the golden path is the easy way.
- Platform as Product: The platform team treats internal developers as customers. Roadmap, user research, SLOs, documentation, onboarding — all the things a good product team does.
- Self-Service: Developers should be able to provision what they need without filing tickets. The platform enables autonomy within guardrails.
Kubernetes Recipes
A practical guide for container orchestration and deployment by Grzegorz Stencel & Luca Berton (Apress).
Watch on Skillshare →Why Regulated Enterprises Need Platform Engineering
- Compliance by default: Security policies, data residency rules, and audit logging built into the platform — not bolted on per team
- Audit efficiency: One platform to audit instead of 50 different team configurations
- Consistent security posture: Every service gets the same encryption, network policies, and access controls
- Faster time to market: Developers focus on business logic, not infrastructure. Typical improvement: 60-80% reduction in time-to-production for new services.
- Cost optimisation: Centralised resource management reduces cloud waste by 30-50%
The Platform Team
A mature platform team typically includes:
- Platform product manager — Owns the roadmap, talks to internal users, prioritises features
- Platform engineers — Build and maintain the IDP (Kubernetes, Crossplane, ArgoCD, Backstage, etc.)
- Developer advocates — Onboard teams, write documentation, gather feedback
- SREs — Ensure platform reliability, define SLOs, manage incidents
Typical ratio: 1 platform engineer per 10-15 application developers is a common starting target.
Optimizing Azure DevTest Labs
Enhance performance, security, and cost efficiency of Azure DevTest Labs.
Start on Pluralsight →