Platform Engineer Roadmap for Beginners
A 14-month path from zero to junior Platform Engineer. Linux, Kubernetes, Terraform, GitOps, Backstage, observability, and building internal developer platforms — no prior experience needed.
What a Platform Engineer does
What is this roadmap and who is it for?
A platform engineer builds the internal infrastructure that software teams use every day — the CI/CD pipelines that test and deploy their code, the Kubernetes clusters that run their services, the developer portals that let them spin up environments without filing a ticket. The job is to reduce the cognitive load on developers, so they can focus on building products instead of managing infrastructure.This roadmap treats the platform as a product — because that's exactly what it is. You're not just configuring tools, you're designing a system that has users, feedback loops, adoption metrics, and a roadmap of its own. That mindset is what separates platform engineers from DevOps engineers who happen to also run a cluster.One thing we want to be upfront about — platform engineering touches Linux, networking, cloud, containers, Kubernetes, Terraform, CI/CD, security, and observability simultaneously. The temptation is to learn everything at once. This roadmap is built to resist that by going deep on one layer at a time before adding the next.
Before you start — 3 Things to Keep in Mind
- 1Linux and networking are the foundation. Everything in platform engineering runs on Linux servers talking over networks — skip this and every tool after it will feel opaque.
- 2The hardest problem in platform engineering isn't technical complexity — it's getting developers to actually use what you build. Start thinking about that from day one.
- 3Build real things at every step. A Terraform module that provisions something you can SSH into teaches more than a week of watching someone else do it.
Estimated duration
This roadmap takes 14 months at a pace of 15 to 20 hours per week.
If you can only commit 10 hours per week, plan for 18 to 22 months.
Consistency matters far more than speed.
Before you begin — what you need
- 1A computer — Windows, Mac, or Linux. WSL2 on Windows gives you a Linux terminal for free.
- 2A free AWS, GCP, or Azure account — free tiers are sufficient for the majority of this roadmap.
- 3A GitHub account — free, and where your entire portfolio will live.
- 4Docker Desktop or Podman — free, for running containers and Kubernetes locally.
- 5VS Code — free, with excellent extensions for Terraform, Kubernetes YAML, and Backstage development.
- 6No prior platform engineering, DevOps, or cloud experience needed — this roadmap starts from zero.
How platform engineering evolved over time.
DevOps and the Collaboration Revolution
The DevOps movement emerged from frustration with the wall between development and operations teams. Agile methods, continuous integration, and tools like Chef and Puppet began closing the gap. But as teams grew and cloud complexity increased, a new problem appeared: developers were spending too much time on infrastructure tasks instead of building products.
Netflix, Google, and the IDP Pioneers
Netflix built Spinnaker for multi-cloud continuous delivery. Google published its Site Reliability Engineering practices. Spotify was building an internal portal to manage hundreds of microservices. These companies weren't doing this for the sake of tools — they were doing it because unmanaged growth in developer teams created chaos, and internal platforms were the solution.
Kubernetes and Terraform: The Platform Stack Takes Shape
Kubernetes was open-sourced by Google in June 2014. Terraform 0.1 shipped in July 2014. Together, these two tools gave platform engineers the primitives they needed: a standard way to run containerised workloads and a standard way to describe cloud infrastructure as code. Almost every internal developer platform built in 2026 runs on top of both.
The IDP Gets a Name and a Framework
Evan Bottcher defined the Internal Developer Platform in 2018 as 'a foundation of self-service APIs, tools, and services as an internal product.' In 2019, Spotify open-sourced Backstage — their internal portal framework — and the CNCF adopted it in 2022. The same year, the book Team Topologies gave platform teams a formal organisational model: a dedicated team that enables other teams by reducing their cognitive load.
Platform Engineering Becomes a Job Title
By 2020, 'platform engineering' appeared consistently in job postings at companies of all sizes. GitOps (using Git as the source of truth for infrastructure state) became standard practice. ArgoCD and Flux matured into production-grade tools. Backstage attracted contributors from hundreds of companies. The field went from niche to expected at any organisation with more than a few software teams.
IDPs Go Mainstream and AI Enters the Platform
Industry surveys report that approximately 55% of organisations have dedicated platform teams in 2025, with Gartner projecting 80% by 2026. Crossplane extended the Kubernetes API to manage cloud resources across providers. Policy-as-code with OPA/Gatekeeper became standard compliance infrastructure. AI-assisted tools began appearing in platforms — generating IaC, triaging alerts, and powering predictive autoscaling. Platform engineering is now a recognised discipline with its own conferences (PlatformCon), community organisations, and career paths.
In 2026, platform engineering is one of the fastest-growing specialisations in DevOps and cloud. Mature organisations report developer-to-platform team ratios around 20:1, effectively doubling developer throughput compared to 5:1 ratios in legacy organisations. Senior platform engineers command $150K to $180K+ in North America. The tools are settled: Kubernetes, Terraform, Backstage, ArgoCD, Prometheus, and OPA form the canonical platform stack. The challenge has shifted from 'how do we build this' to 'how do we get developers to love using it'.
What's shaping platform engineering in 2026.
IDPs Are Now an Organisational Expectation
Approximately 55% of organisations have dedicated platform teams in 2025, with Gartner projecting 80% by 2026. The question is no longer 'should we have a platform team?' but 'how mature is our IDP and how do we improve it?' This creates a significant and growing demand for engineers who understand how to build and operate internal platforms.
Backstage Is the De Facto Developer Portal
Backstage, now a CNCF incubating project, has become the standard framework for internal developer portals. Most organisations either use Backstage directly or use a Backstage-compatible commercial alternative (Port, Cortex, Roadie). Understanding how to install, configure, extend, and run Backstage is a baseline skill for platform engineering in 2026.
GitOps Is the Default Deployment Model
Git is the source of truth for both application manifests and infrastructure state. ArgoCD and Flux automatically reconcile cluster state with what's in the repository. Every deployment is a Git commit, every rollback is a revert, and every change is auditable. Platform engineers design the GitOps workflows that development teams use — not just the clusters they run on.
Security and Policy Are Delivered Through the Platform
Policy-as-code with OPA/Gatekeeper enforces security rules at admission time — before a workload can run, not after. Container image scanning, secret detection, and compliance checks are CI/CD pipeline steps, not manual reviews. Platform engineers build the guardrails that make developers productive without making the organisation vulnerable.
AI Is an Amplifier in the Platform Stack
AI tools now assist platform engineers with IaC generation, log analysis, alert triage, and predictive autoscaling. LLM-powered chatbots inside developer portals help teams find documentation and understand platform capabilities. AI accelerates the work but doesn't design the systems — the architecture, reliability, and governance decisions remain human.
The honest state of platform engineering jobs in 2026.
What's happening in the market
High Demand and Strong Compensation
Gartner projects 80% of organisations will have platform teams by 2026. Senior platform engineers command $150K to $180K+ in North America, above comparable DevOps and SRE roles. The developer-to-platform team ratio averages around 20:1 in mature organisations, meaning each platform engineer multiplies the productivity of 20 developers — a value proposition that's easy for engineering leadership to justify.
Remote and Hybrid Work Are Standard
Platform work is cloud-centric and asynchronous by nature — teams are distributed, infrastructure is remote, and collaboration happens in GitHub and Slack. Remote platform engineering roles are common at every company size, and the global market means your location is a smaller factor than your portfolio and demonstrated skills.
The Field Is Democratising
Platform engineering is no longer only for senior architects. Mid-career DevOps and SRE engineers are transitioning into platform roles, and junior roles are emerging at companies with established platform practices. Specialisations are appearing — Developer Experience Engineer, Security Platform Engineer, Platform Reliability Engineer — each with a narrower and more accessible entry point.
The Hardest Problem Is Adoption, Not Technology
The number one challenge in platform engineering is getting developers to use the platform. A technically excellent platform that no one adopts is a failure. Platform engineers who understand developer experience, gather feedback systematically, and treat adoption as a product metric are significantly more effective — and more hireable — than those who only think about the infrastructure layer.
What you can do instead — or as well
Site Reliability Engineering (SRE)
SRE focuses on the reliability and availability of production systems — SLOs, error budgets, incident response, and capacity planning. Platform engineering and SRE overlap significantly in tooling; the difference is focus. SRE focuses on system behaviour under load. Platform engineering focuses on the infrastructure developers use to build and deploy. Many platform engineers started in SRE.
Cloud Architecture
Cloud architects design multi-cloud strategies, cost optimisation frameworks, and reference architectures. The platform engineering foundation — Terraform, Kubernetes, networking, IAM — is the direct path into cloud architecture. The shift is from building the platform to designing the strategic patterns that the platform implements.
Developer Experience (DevEx) Engineering
A specialisation focused on the usability and productivity of developer tools — internal documentation, API design, developer portals, tooling UX. Backstage development, developer portal product management, and DevEx metrics are the domain. It's platform engineering with a product and UX lens, and it's one of the fastest-growing specialisations in the field.
Freelance Platform Consulting
Small and mid-size companies need help setting up Kubernetes, building CI/CD pipelines, migrating to cloud, and deploying Backstage. Freelance platform engineering is a real path — platform engineers who can help a 50-person company go from manual deployments to a working IDP in 6 weeks are genuinely valuable and consistently in demand.
Platform Product Management
As IDPs mature, they need product managers who understand both developer needs and infrastructure constraints. A platform engineer who develops strong product instincts — roadmap planning, adoption metrics, stakeholder communication — is well positioned to move into this hybrid role, which is rare and compensated accordingly.
Platform engineering is one of the highest-leverage roles in a software organisation — you multiply the productivity of every development team you serve. The 14-month timeline reflects the real breadth of what the role requires. Engineers who rush through it produce platforms that look complete and break under the weight of real usage patterns.
Your step-by-step guide.
Foundation
The ground everything else stands on
3 steps
Core Skills
The must-have tools of the job
4 steps
Advanced
What separates beginners from job-ready developers
4 steps
Professional
The layer that makes you hireable
3 steps
A simple 14-month learning path.
Linux and Networking Foundations
Terminal navigation, file permissions, SSH, Bash scripting, IP addressing, DNS, HTTP/S, diagnostic tools
Git, Scripting, and Containers
Git workflow, Python and Bash scripting, YAML and JSON, Dockerfile fundamentals, Docker Compose, image registries
Container Depth and Cloud Basics
Multi-stage builds, image scanning, container registries, cloud compute and networking fundamentals, IAM basics, cloud CLI
Kubernetes Foundations
Pods, Deployments, Services, ConfigMaps, Secrets, kubectl, YAML manifests, Minikube, namespaces
Kubernetes Advanced and Terraform
Ingress, HPA, PVCs, resource limits, Terraform providers, state, modules, remote backend, full cloud network
CI/CD Pipelines
GitHub Actions workflows, reusable workflows, pipeline stages, image tagging, secrets in CI, AWS OIDC
GitOps and Helm
ArgoCD installation and app management, Flux basics, Helm charts and values, Kustomize overlays, platform golden path chart
Backstage Developer Portal — Part 1
Backstage installation, Software Catalog, catalog-info.yaml, TechDocs, GitHub integration, entity types
Backstage Developer Portal — Part 2
Scaffolder templates, Software Templates, plugin installation (ArgoCD, Kubernetes), portal deployment on Kubernetes, adoption metrics
Observability
kube-prometheus-stack, Prometheus ServiceMonitors, Grafana dashboards, Alertmanager, Loki log aggregation, OpenTelemetry basics
Security, Policy, and Secrets
Kubernetes RBAC, OPA Gatekeeper policies, image scanning in CI, external-secrets-operator, HashiCorp Vault basics, network policies
Self-Service Infrastructure with Crossplane
Crossplane providers, Composite Resources, Claims, Compositions, Backstage Scaffolder integration, FinOps with Kubecost
Developer Experience and Multi-Cloud
Platform-as-product mindset, DORA metrics, developer surveys, multi-cluster ArgoCD, service mesh overview, platform disaster recovery
Portfolio and Interview Preparation
Complete IDP project, architecture diagrams, case study writing, CKA exam prep, system design practice, community contributions
What to focus on first.
Linux and Networking
Every platform tool runs on Linux servers communicating over networks. Engineers who can't navigate the terminal or debug a connectivity issue between services will struggle with every other layer in this roadmap.
Git and Scripting
Platform infrastructure is code. Every Terraform config, every Kubernetes manifest, every Backstage template lives in Git. Python and Bash scripting turn manual tasks into automation. Both are used daily from the first month of the job.
Docker and Containers
Everything a platform team manages runs in containers. Kubernetes, CI/CD, Backstage itself — all containers. Docker is the entry point, and understanding images, layers, and registries makes Kubernetes significantly more intuitive.
Kubernetes
Kubernetes is the runtime layer of virtually every internal developer platform. Pods, Deployments, Services, RBAC, and resource management are the vocabulary of daily platform work. Most platform engineering job interviews test Kubernetes extensively.
Terraform
Platform infrastructure that isn't in code can't be reproduced, reviewed, or managed at team scale. Terraform is the most widely used IaC tool and the standard for provisioning the clusters, networks, and databases that IDPs run on.
CI/CD Pipelines
Platform engineers design the pipeline patterns that every development team uses. Understanding how to build, structure, and secure CI/CD workflows — including reusable workflows — is central to the platform team's value proposition.
GitOps and Helm
GitOps is the default Kubernetes delivery model. ArgoCD, Helm charts, and Kustomize overlays are how platform teams ship consistent deployments across multiple environments and multiple teams without duplicating configuration.
Backstage
Backstage is the developer-facing surface of the platform — the interface where developers find services, trigger pipelines, read documentation, and request resources. Understanding how to install, extend, and adopt it is a distinguishing skill for platform engineers in 2026.
Observability
Platform teams deliver observability as a service. Every development team should be able to see their application's health metrics and logs without configuring a single scrape job. Prometheus, Grafana, and the kube-prometheus-stack are the platform's monitoring baseline.
Security and Policy
Platform teams set the guardrails that protect the entire organisation's cloud environment. OPA/Gatekeeper, RBAC, image scanning, and secrets management are what prevent individual teams from making security mistakes that affect everyone else.
Crossplane and Self-Service
The highest-leverage capability a platform can provide is genuine self-service — a developer who can provision a database through the portal without filing a ticket. Crossplane and well-designed Backstage templates are how mature platforms deliver this.
Developer Experience and Portfolio
The best platform is one developers love to use. Adoption metrics, developer surveys, and product thinking are what ensure the platform actually reduces cognitive load. And a documented, working IDP is the only thing that proves you can do all of the above together.
Problems every beginner faces — and how to get through them.
The Breadth Is Paralysing
What it looks like
You look at a platform engineering job description — Kubernetes, Terraform, ArgoCD, Backstage, OPA, Prometheus, Crossplane, Vault — and feel like you'd need five years just to know what half of these things are.
How to get through it
Follow this roadmap's sequence without skipping. Linux, containers, Kubernetes, Terraform — in that order, with real projects. Each layer makes the next one significantly more learnable. Engineers who jump straight to Backstage without Kubernetes knowledge spend three months confused about why a deployment isn't working. Engineers who build from the foundation spend three months actually building.
Developer Adoption Is Harder Than the Technology
What it looks like
You build a beautiful Backstage portal with full catalog coverage and golden path templates. Three months later, developers still file Jira tickets to request databases instead of using the self-service form you spent a month building.
How to get through it
Treat adoption as a product problem, not a communication problem. Talk to the developers who aren't using the platform — not to convince them, but to understand why. Often the form is too complex, the error messages are confusing, or the template outputs don't match what they actually need. Fix the friction, don't market around it.
Cloud Bills for Platform Experiments
What it looks like
You spin up a managed Kubernetes cluster for practice, add a NAT gateway, a load balancer, and three node groups, and receive a $200 bill at the end of the month.
How to get through it
Use Minikube or Kind for local Kubernetes practice — they're free and sufficient for everything in the foundation and core months. When you do use cloud clusters, use the smallest possible node sizes, destroy clusters after each session, and set billing alerts at $10 before starting any cloud work. EKS on one t3.medium node with a one-week TTL costs pennies.
Platform Breaks Break Everyone
What it looks like
You push a change to a Gatekeeper policy that inadvertently prevents any new pod from being scheduled. Five development teams are affected before you notice.
How to get through it
Test policy changes in a dedicated canary namespace before applying them cluster-wide. Use OPA Gatekeeper's warn mode instead of deny mode when introducing new policies. Maintain a rollback plan for every platform change, and communicate changes to development teams with enough lead time to prepare. Platform changes have blast radius — treat them like production deployments.
Kubernetes Confusion Before It Clicks
What it looks like
You read the Kubernetes documentation and still can't form a clear mental model of what a pod is, why there are six resource types, or why your service isn't routing to your deployment.
How to get through it
Deploy one application end-to-end on Minikube before reading any architecture documentation. kubectl describe and kubectl logs answer most questions — run both before changing anything. The concepts click from use, not from reading. Accept a month of confusion as the price of admission to a skill that compounds for years.
Imposter Syndrome in a Field That Touches Everything
What it looks like
Platform engineering requires Linux, networking, cloud, containers, Kubernetes, Terraform, CI/CD, security, observability, and product thinking. You feel like you'll never know enough.
How to get through it
Nobody knows all of it. Even senior platform engineers google Terraform provider documentation and ask Slack about Kubernetes scheduler behaviour. The skill is knowing how to learn the next tool when you need it — which is exactly what this roadmap builds. Focus on one layer at a time, finish real projects, and let the breadth accumulate naturally.
Can't Get the First Platform Role
What it looks like
Junior platform engineering postings are rare — most ask for 3 years of Kubernetes and IaC experience. You feel like the entry point doesn't exist.
How to get through it
Build the minimal complete IDP from this roadmap and put it on GitHub with a clear README and architecture diagram. Many people enter platform engineering via DevOps or infrastructure roles and pivot once they have cluster management and Terraform experience. The CKA certification signals real Kubernetes competence without requiring prior employment. And platform consulting for smaller companies — helping them set up their first CI/CD pipeline or Kubernetes cluster — builds the exact experience hiring managers are looking for.
You're ready for a junior Platform Engineer role when you can….
Navigate and administer a Linux server entirely from the command line, debug a service that's failing to start, and explain why it's failing from the logs.
Deploy an application to a Kubernetes cluster using YAML manifests and kubectl, configure RBAC for namespace isolation, and debug a failing pod from describe and logs.
Write Terraform that provisions a complete Kubernetes cluster on a cloud provider — apply, verify, and destroy it reproducibly — and explain every resource in the configuration.
Build a GitHub Actions pipeline that builds a Docker image, pushes it to a registry, and triggers an ArgoCD sync to update a running Kubernetes Deployment.
Install and configure Backstage, register 5 services in the Software Catalog with TechDocs, and create a Scaffolder template that automates new service creation.
Deploy OPA Gatekeeper to a cluster, enforce at least two admission policies, and explain what each one prevents and why it matters.
Deploy Prometheus and Grafana using the kube-prometheus-stack, add a ServiceMonitor for a custom application, and configure an alert rule that fires to Slack.
A good platform engineer isn't someone who knows every CNCF project. They understand the problem each layer solves, can build a working IDP from scratch, measure its adoption, and treat the developer experience as seriously as the infrastructure reliability. Fourteen months is a real investment — and the working platform you finish with is proof that you can do the actual job.
You now have a clear path forward.
Platform engineering compounds the same way other infrastructure skills do — every IDP you build teaches you something the next one benefits from, and every time a development team tells you your platform saved them an hour of work, the motivation compounds too. The roadmap gives you the order. The depth comes from building real platforms, measuring their adoption, and iterating on what developers actually need.
The goal was never to learn every CNCF project or certify in every cloud. It was to reach a point where you can look at a software team buried in infrastructure toil, design a self-service platform that removes that toil, build it in code, and measure whether it worked.
Start with Linux, SSH into your first server, and keep going from there.
No login required to share feedback
Frequently Asked Questions.
Trusted places to keep learning.
Backstage Documentation — backstage.io/docs
The official Backstage documentation — getting started guides, Software Catalog configuration, Scaffolder templates, plugin development, and deployment guides. The most authoritative reference for the developer portal layer of this roadmap. The community section also links to hundreds of open-source plugins.
Kubernetes Official Documentation
The Kubernetes official docs — comprehensive guides on every resource type, the scheduler, networking, storage, RBAC, and the API. The Concepts section builds the mental model; the Tasks section shows you how to do specific things. Also hosts the Kubernetes tutorials and the free interactive playground.
HashiCorp Terraform Tutorials
HashiCorp's step-by-step Terraform tutorials for AWS, Azure, and GCP. Each tutorial provisions real infrastructure and teaches the Terraform workflow from the ground up. The most structured free resource for infrastructure as code on this roadmap.
Internal Developer Platform — internaldeveloperplatform.org
The community hub for platform engineering — articles, the annual State of Platform Engineering report, IDP architecture guides, and a curated tooling landscape. The best single place to understand what mature IDPs look like and what problems they solve, beyond the specific tools this roadmap covers.
CNCF Landscape and Projects
The Cloud Native Computing Foundation's interactive landscape of every cloud-native tool and project. An invaluable reference for understanding where each tool (ArgoCD, Backstage, Crossplane, OPA, Prometheus) sits in the broader ecosystem, and what alternatives exist in each category.
Keep going
Ready to go further?
Explore the Resource Hub for practical guides, honest reviews, and quick-reference cheatsheets designed to help you build faster.