Hands-on labs for AIOps and MLOps

Learn to run AI in production by running it.

Every lab gives you your own cloud machine with Kubernetes ready to go, a short explainer for each step, and automatic checks that confirm you got it right. Nothing to install.

Early access opens soon. We'll email you when your spot is ready.

Every lab works the same way

01

Get the idea

A short video or reading explains the concept before you touch anything.

02

Do it for real

Work in a browser terminal and editor connected to your own machine, with root access.

03

Check your work

Run the check for each step. Pass it and the next step unlocks.

The first labs

In development now. People on the waitlist get them first.

LLMOps

Serve a model with vLLM

Deploy an inference server on Kubernetes, load test it, and read its metrics.

LLMOps

Scale inference with llm-d

Route across replicas and compare smart scheduling with plain load balancing.

AIOps

Build an AI SRE agent

Give an agent read-only tools and have it diagnose a cluster we broke on purpose.

LLMOps

LLM observability with OpenTelemetry

Trace requests, tokens, latency, and cost for an LLM app.

MLOps

Evaluation as a release gate

Score model outputs automatically and block a bad change before it ships.

MLOps

A model pipeline, end to end

Track, register, deploy, and monitor a model with MLflow and KServe.

Built for engineers who learn by doing

Break things safely

Your lab machine is yours alone. Wreck it, reset it, and try again.

The tools you'll use at work

Kubernetes, Prometheus, MLflow, and the model servers teams run in production.

Short and focused

Each lab covers one skill and fits in a single sitting.

Be first into the lab

Join the waitlist and we'll email you when early access opens.