Vibe Engines is free, interactive AI engineering & system design — runs in your browser. Pick a learning path →
18 stations. 3 tracks. From Linux and containers to Kubernetes, infrastructure-as-code and CI/CD, then internal developer platforms, golden paths and SRE — become the engineer who builds the platform everyone else ships on.
18 stations on 3 ladders, in order — each rung is written assuming the one above it. Mark them off as you go; your progress is saved in this browser.
6 rungs · ~6h · ends at Programming for Platforms.
F1 · Beginner · 55 min
Everything a platform runs on is Linux, and the shell is where you live. Learn the filesystem, processes and signals, permissions, systemd, and enough shell scripting to automate the boring things — because a platform engineer who fears the terminal is a platform engineer who copy-pastes runbooks.
Build itA service is eating memory on a box. Find the process, read its limits, and restart it cleanly — all from the shell.
✓ CheckpointExplain what a signal does to a process, and why kill -9 is the one your service never gets to handle.
F2 · Beginner · 55 min
Distributed systems are networking problems wearing trench coats. Learn TCP/IP, DNS, HTTP, load balancing and TLS, plus the cloud-native additions — service discovery, ingress, and why "it works locally but not in the cluster" is almost always a networking answer.
Build itA pod can’t reach a database by name. Walk the resolution path — DNS, service, network policy — and find where it breaks.
✓ CheckpointTrace a request from an ingress to a pod and name every hop. Any gap you cannot narrate is one you cannot debug.
F3 · Beginner · 60 min
Containers are the unit a platform ships. Learn what a container actually is (namespaces + cgroups, not a VM), how image layers and caching work, writing Dockerfiles that build fast and ship small, and registries — the foundation the entire orchestration stack sits on.
Build itCut a 1.2 GB image to under 100 MB with a multi-stage build. What did the fat version ship that production never needed?
✓ CheckpointExplain why Docker layer order changes build time, and which line invalidates your cache most often.
F4 · Intermediate · 60 min
Platforms are built on cloud primitives. Learn compute, storage and networking building blocks, IAM and the shared-responsibility model, managed services vs self-hosted trade-offs, and the cost awareness that separates a platform engineer from a big cloud bill.
Build itChoose between a managed database and self-hosting one on Kubernetes. List the trade-offs your team will actually feel.
✓ CheckpointDraw the shared responsibility line for a managed service, and name what you wrongly assumed the provider handles.
F5 · Beginner · 45 min
GitOps means Git is the platform’s source of truth, so you must understand it deeply. Learn Git’s object model, branching strategies, and why declarative infrastructure lives in repositories — the substrate every modern deployment workflow is built on.
Build itExplain why "the cluster state should equal what’s in Git" is the whole idea of GitOps — and what breaks when it isn’t.
✓ CheckpointExplain when rebase is right and when it rewrites history someone else has already pulled.
F6 · Intermediate · 55 min
Platform engineering is software engineering aimed at infrastructure. Learn Go or Python well enough to write operators, controllers and CLIs, work with APIs and YAML/JSON, and treat your platform as a product with code — not a pile of shell scripts nobody dares touch.
Build itWrite a small script that reconciles desired vs actual state for a resource. You just built the core idea of a controller.
✓ CheckpointExplain what makes a platform script idempotent, and why that matters more than whether it is short.
6 rungs · ~6h · ends at Platform Security.
T1 · Intermediate · 70 min
Kubernetes is the platform engineer’s operating system. Learn pods, deployments, services and the control loop that makes it all work — desired state reconciled continuously. Master kubectl, understand the scheduler, and grasp why "declare what you want, let the controller converge" is the whole philosophy.
Build itA deployment won’t roll out. Use describe, events and logs to diagnose it — the debugging loop you’ll run a thousand times.
✓ CheckpointExplain what the control loop does when you delete a pod by hand, and why that is the whole idea.
T2 · Advanced · 70 min
Past the basics lies the real platform. Learn ConfigMaps and Secrets, RBAC, network policies, storage and StatefulSets, resource requests/limits, and Helm or Kustomize for packaging — the pieces that turn "runs a container" into "runs a company’s workloads safely".
Build itPackage an app with Helm so any team can deploy it with three values. What do you expose, and what do you lock down?
✓ CheckpointExplain what a readiness probe changes that a liveness probe does not, and what breaks when you swap them.
T3 · Advanced · 65 min
Clicking in a console doesn’t scale and can’t be reviewed. Learn Terraform (or OpenTofu/Pulumi): declarative infrastructure, state management and its dangers, modules for reuse, and the discipline that makes your entire cloud reproducible from a repository and a plan.
Build itTwo engineers run terraform apply at once and corrupt state. Explain what happened and how remote state + locking prevents it.
✓ CheckpointExplain what Terraform state protects, and what happens when two engineers apply at once.
T4 · Advanced · 65 min
The platform’s job is to move code to production safely and often. Learn pipeline design, build/test/deploy stages, GitOps with Argo CD or Flux (the cluster pulls desired state from Git), and progressive delivery — canary and blue-green — so deploys are boring instead of terrifying.
Build itDesign the deploy path where a bad release auto-rolls-back on failing health checks. What signal triggers the rollback, and when?
✓ CheckpointExplain what GitOps reconciliation does when someone changes the cluster by hand.
T5 · Advanced · 60 min
You can’t operate what you can’t see. Learn the three pillars — metrics, logs, traces — plus Prometheus and Grafana, structured logging, distributed tracing, and how to build dashboards and alerts that answer "is it broken and why?" without drowning on-call in noise.
Build itA request is slow across five services. Which pillar tells you where the time went — and what must be instrumented for it to work?
✓ CheckpointExplain which of metrics, logs and traces answers “is it broken”, “why”, and “where”.
T6 · Advanced · 55 min
The platform is the blast radius for everyone on it. Learn secrets management, image scanning and supply-chain security, policy-as-code (OPA/Kyverno), and the paved-road principle: make the secure way the easy way, so every team gets security by default without becoming security experts.
Build itWrite a policy that blocks any pod running as root from being admitted. Why enforce it at the platform, not in each team’s YAML?
✓ CheckpointExplain why a pod running as root is a risk even inside a private cluster.
6 rungs · ~6h · ends at Platform as a Product.
P1 · Advanced · 60 min
SRE is the discipline of running things reliably with engineering, not heroics. Learn SLIs, SLOs and error budgets, the toil-reduction mandate, capacity planning, and the culture that treats reliability as a feature you can measure, budget and trade against velocity.
Build itSet an SLO for an API and the burn-rate alerts around it. When does the error budget say "stop shipping features and fix reliability"?
✓ CheckpointExplain what an error budget lets you say no to — and what it lets you say yes to.
P2 · Advanced · 60 min
The frontier of the discipline: building the platform product itself. Learn golden paths, self-service provisioning, platform portals (Backstage), and the paved-road philosophy — abstracting Kubernetes complexity so a product engineer ships a service in an afternoon, not a fortnight.
Build itDesign the "new service" button: what does a product team fill in, and what does the platform generate — repo, pipeline, infra, dashboards?
✓ CheckpointExplain why a platform nobody adopts is a failed platform, however good the technology.
P3 · Advanced · 55 min
Everything breaks; what matters is how you respond. Learn on-call that doesn’t burn people out, incident command, effective runbooks, and blameless post-mortems that turn outages into permanent fixes instead of repeated 2am pages and quiet resentment.
Build itDesign an on-call rotation for a 6-person team that is humane AND covers 24/7. What automation removes the pages nobody should get?
✓ CheckpointExplain why alerting on symptoms beats alerting on causes, and what a 3am page should mean.
P4 · Intermediate · 50 min
A platform that ignores cost gets a very unpleasant meeting. Learn cloud cost visibility, right-sizing, autoscaling, spot instances, and FinOps practices that give teams the cost feedback to make good trade-offs — because the cheapest architecture nobody can operate is not actually cheap.
Build itYour cluster runs at 20% utilization "for safety". Design the autoscaling that keeps headroom without paying for idle nodes all night.
✓ CheckpointExplain how idle resources accumulate on a platform and who is incentivised to notice.
P5 · Advanced · 55 min
One cluster becomes many: regions, environments, tenants. Learn multi-cluster patterns, service meshes (Istio/Linkerd) for traffic and mTLS, multi-tenancy isolation, and the disaster-recovery and failover design that keeps a global platform up when a region goes dark.
Build itA whole region fails. Trace what keeps serving and what doesn’t under your failover design — and what data you might lose.
✓ CheckpointExplain what multi-cluster buys you and the operational cost you just took on.
P6 · Intermediate · 45 min
The mindset shift that defines senior platform engineers: your users are developers, and the platform is a product they choose to use. Learn to gather feedback, measure adoption and developer experience, treat internal teams as customers, and grow from running infrastructure to shaping how a whole org ships.
Build itTeams keep bypassing your platform for their own scripts. Diagnose it as a product problem — what does your "product" get wrong?
✓ CheckpointName your platform's users and the last thing you shipped because they asked, rather than because it was interesting.
The common questions before you start — how long it takes, whether to follow it in order, and how it stays current.
It runs 18 stations across 3 tracks — roughly ~17h of focused learning, plus the time you spend actually building. It is self-paced, so most people work through it over a few weeks, an evening or a single station at a time.
The tracks are ordered so each station builds on the one before, and following them start to finish is the intended path. But every station also stands alone — if you already have the foundations, jump straight to the part you need.
Yes. The whole roadmap, and every handbook, lab, and challenge it links to, is free and open — no sign-up and no paywall.
It teaches the durable fundamentals of the role first, then the tooling and the AI-era shifts on top — so most of it stays relevant as individual tools churn, and it is revised as the role itself changes.
Anyone stepping into or leveling up in the Platform Engineer role — whether you are switching in, early-career, or a senior filling gaps. Start where you are; the ladder shows what is left.
See this alongside everything else on the same subject — handbooks, system designs, challenges and tools, in one place.