All Roadmaps
Technology Roadmap
Platform Engineer Roadmap
Complete platform engineering roadmap. Build Internal Developer Platforms (IDPs), golden paths, self-service infrastructure, and the developer experience tools that 10x engineering velocity.
6–10 months
7 phases
FoundationIntermediateAdvancedExpert
Phase 1
DevOps & Cloud Foundations
Platform engineers build on DevOps — master it first
Foundation4–6 weeks
What to learn
- Linux, shell scripting, Git workflows
- Docker — images, multi-stage builds, registries
- AWS/GCP core services — compute, storage, VPC, IAM
- Terraform basics — resources, modules, state
- CI/CD pipeline concepts — build, test, deploy
- Kubernetes fundamentals — pods, deployments, services
Key tools
dockerTerraformAWS CLIkubectlgit
Phase 2
Kubernetes as a Platform
Kubernetes IS the platform foundation for most companies
Intermediate4–5 weeks
What to learn
- Multi-tenancy — namespaces, vCluster, Capsule, Loft
- RBAC and admission controllers at scale
- Cluster API and cluster lifecycle management
- Service mesh — Cilium, Istio for platform-wide observability
- Storage — StorageClasses, CSI drivers, Velero backup
- Multi-cluster strategies — hub-spoke, fleet management
Key tools
HelmCrossplanevClusterCluster APICilium
Phase 3
Internal Developer Platform Concepts
Understand what an IDP is and why it exists
Intermediate2–3 weeks
What to learn
- Platform vs toolchain — abstraction over infrastructure
- Cognitive load — why IDPs reduce developer toil
- Team Topologies — platform team as internal product team
- CNCF platforms landscape — Backstage, Port, Humanitec
- Build vs buy decisions — open-source vs commercial IDPs
- Platform as a product — roadmapping, user research, adoption
Key tools
BackstagePortHumanitecCortex
Phase 4
Backstage IDP
The open-source platform used by Spotify, Netflix, and 1000+ teams
Intermediate3–4 weeks
What to learn
- Backstage architecture — core, plugins, software catalog
- Software catalog — all services, teams, and ownership in one place
- TechDocs — documentation as code, auto-generated
- Software templates — one-click new service creation
- Custom plugins — React frontend + Node.js backend
- Deploying Backstage on Kubernetes with Helm + PostgreSQL
Key tools
BackstageReactNode.jsPostgreSQLHelm
Phase 5
GitOps & Self-Service Infrastructure
Developers provision their own infra safely via Git
Advanced3–4 weeks
What to learn
- ArgoCD ApplicationSets for multi-team GitOps at scale
- Crossplane — cloud resources as Kubernetes CRDs
- Flux — GitOps for cluster and app configuration
- Policy enforcement — OPA, Kyverno as guardrails
- PR-driven infra changes — plan, review, apply in CI
- Secrets management — External Secrets Operator + Vault
Key tools
ArgoCDCrossplaneFluxKyvernoExternal Secrets Operator
Phase 6
Golden Paths & Developer Experience
Paved roads that make the right thing the easy thing
Advanced3–4 weeks
What to learn
- Defining golden paths — opinionated, supported defaults
- Service templates — deploy new service in one PR
- Deployment abstractions — developers use Git, not kubectl
- Observability by default — auto-instrumented from template
- Secret self-service — Vault + External Secrets integrated
- Developer portal — service catalog, runbooks, ownership
Key tools
Backstage TemplatesVaultExternal Secrets OperatorArgoCD
Phase 7
Platform Metrics & Maturity
Measure platform impact — adoption, reliability, velocity
ExpertOngoing
What to learn
- DORA metrics — deploy frequency, lead time as platform KPIs
- Platform adoption — % teams using golden paths
- Cognitive load reduction — time-to-prod for new services
- Platform SLOs — the platform must be as reliable as prod
- Developer satisfaction — NPS surveys, friction tracking
- Platform roadmapping — product management for internal tools
Key tools
GrafanaLinearBCortexBackstage analytics
Interview Prep
DevOps Interview Prep Bundle — 1000+ Q&A
Every topic on this roadmap has interview questions in the bundle — Docker, Kubernetes, AWS, CI/CD, Linux, SRE, FinOps, System Design. Grab it before your next interview.
Frequently Asked Questions
Common questions about the Platform Engineer roadmap
1What does a platform engineer do?
Platform engineers build Internal Developer Platforms (IDPs) — the self-service tools, golden paths, and infrastructure abstractions that let development teams deploy and operate services without DevOps bottlenecks. They treat the platform as a product, with internal developers as customers.
2How is platform engineering different from DevOps?
DevOps embeds ops knowledge into dev teams. Platform engineering centralizes that knowledge into a dedicated platform team that builds self-service tooling. Platform engineers reduce cognitive load on developers by abstracting Kubernetes, Terraform, and CI/CD into simple, opinionated workflows.
3How long does it take to become a platform engineer?
Platform engineering requires strong DevOps foundations first. Plan 6–10 months: 4–6 months mastering Kubernetes, Terraform, and CI/CD; then 3–4 months learning Backstage, Crossplane, and GitOps patterns for building IDPs.
4What is Backstage and why do platform engineers use it?
Backstage is an open-source Internal Developer Platform framework by Spotify. It provides a software catalog (track all services), TechDocs (docs as code), software templates (scaffold new services), and a plugin ecosystem. Used by Netflix, Spotify, American Airlines, and 1000+ companies.
5What is a golden path in platform engineering?
A golden path is an opinionated, supported-by-default workflow for common tasks — like creating a new microservice, deploying to production, or rotating secrets. It abstracts complexity and guides developers toward best practices. Not the only way, but the recommended way.