Lesson 1 · Free · 2 min
$ kubectl get localqueues -n team-search
Watch lesson 1 freeWhy AI needs a paved road · 2 min

Platform engineering for AI workloads

Cloud · 27 lessons · 1 h 1 min · Advanced

Build the golden paths, GitOps delivery, self-service infrastructure, observability and cost controls that product teams need to ship AI features. You will extend one platform across the whole course and finish by diagnosing a real-shaped failure.

What you'll be able to do

  • Design a golden path as a product, with named customers and measures
  • Set up declarative delivery and promotion with a GitOps reconciler
  • Offer self-service infrastructure with guardrails enforced as policy
  • Queue, share and quota scarce GPU capacity fairly across teams
  • Define SLOs and signals for inference and roll out model versions safely
  • Allocate GPU cost to teams and measure the platform with DORA metrics

How the lessons work

Lessons that move

Narrated, animated lessons with real terminals and code on screen.

Pause and ask

Stuck? Ask out loud or type. The answer plays as a short video made for your question.

Check yourself

Short quizzes inside the lessons show whether it stuck.

Course content

27 lessons · 1 h 1 min · 2 lessons are free
Unit 1 · Platform as a product4 lessons · 9 min
01Why AI needs a paved road2 minFree02Anatomy of a golden path2 minFree
03Know your customers2 min
04Build, buy or skip3 min
Unit 2 · GitOps and delivery4 lessons · 9 min
05Desired state and reconcilers2 min
06Argo CD or Flux2 min
07Promotion across environments3 min
08Making sync safe3 min
Unit 3 · Self-service and guardrails3 lessons · 7 min
09Templates and scaffolding3 min
10Platform APIs with Crossplane2 min
11Guardrails as policy2 min
Unit 4 · AI workloads on the platform7 lessons · 16 min
12What is different about AI workloads2 min
13Requesting GPUs2 min
14Queueing with Kueue2 min
15Fair sharing and borrowing2 min
16Sharing one GPU3 min
17Dynamic Resource Allocation2 min
18Rolling out model versions2 min
Unit 5 · Observability for AI3 lessons · 7 min
19Instrumenting inference2 min
20GPU and queue signals2 min
21SLOs and burn rates2 min
Unit 6 · Governance and cost2 lessons · 4 min
22Allocating GPU cost2 min
23Unit economics for AI2 min
Unit 7 · Measuring the platform, and the capstone4 lessons · 9 min
24Measuring the platform2 min
25What the research says about AI and delivery2 min
26Capstone: the launch2 min
27Capstone: the diagnosis2 min