Skip to content

Most organisations start with a pilot: one course or one team. Request a pilot

RolesRole: IT operations / site reliability

Practise your first on-call incident before it's real

Lead an outage from the first alert to the all-clear: investigate, roll back, communicate and write it up, in a simulated company where nothing real breaks.

Fig. 1IT operations / SRE
Runs in
A web browser, 3D included
One exercise
20 to 55 minutes
Practice levels
Six, from worked example to exam
Working tools
3 used in these exercises

A real exercise

Checkout errors jump right after a release. Dashboards, logs and the deploy history point in different directions. You find the flag that broke payments, switch it off without switching off the rest, and tell support what to say to customers.

01

How it works

Every exercise follows the same rhythm as the real job: a problem arrives, you investigate, you act, and the simulated company responds.

  1. 01

    You're paged

    Alerts fire and customers can't pay.

  2. 02

    Investigate

    Dashboards, logs, the release pipeline and the people who own the service.

  3. 03

    Mitigate

    Roll back or switch off the broken feature, safely.

  4. 04

    Communicate and review

    Updates during the incident and a clear timeline afterwards.

02

What you practise

Concrete tasks taken from the exercises on this page. Each one is done with real data, under real constraints, inside the same simulated company.

  • T01

    Incident command

    Decide priorities, assign work and keep the response organised.

  • T02

    Release forensics

    Find which change caused the problem.

  • T03

    Feature flags and rollbacks

    Reduce the damage without causing more.

  • T04

    Stakeholder updates

    Plain, accurate updates for non-engineers.

What you get

Experience you can point to.

  1. 01

    Calm under pressure

    You've been through it before, so the real page is less frightening.

  2. 02

    A method

    A repeatable way to work through any incident.

  3. 03

    Interview examples

    Concrete incidents you handled and can talk through.

03

The tools you work with

Working tools, not screenshots. They read the same live data as the rest of the simulated company, and what you change in one shows up in the others.

  • 01

    Incident console

    The live state of services and hosts during an outage, with the runbook actions.

  • 02

    Feature flags

    The switches a service reads, and which one is breaking it.

  • 03

    Data pipelines

    The pipelines feeding production, and which one is sending bad data.

04

Six levels, one exercise

Start by watching it done, then take the hints away one step at a time until you can do it alone.

  1. L1

    Watch it done

    The exercise solved step by step, with what each step changed.

    Hints
    Read only
    Graded
    No
  2. L2

    Step by step

    The next move is pointed out, every hint is open, goals update as you work.

    Hints
    All hints
    Graded
    Yes
  3. L3

    With hints

    Every hint is open, but finding the next step is up to you.

    Hints
    All hints
    Graded
    Yes
  4. L4

    On your own

    Hints stop short of the worked answer. Goals still update as you go.

    Hints
    Some hints
    Graded
    Yes
  5. L5

    Exam

    No hints and no AI helper. Results appear only after you submit.

    Hints
    None
    Graded
    Yes
  6. L6

    Free play

    An open what-if sandbox. Try anything; nothing counts towards a grade.

    Hints
    All hints
    Graded
    No
05

A debrief after every attempt

You see what went well, what did not, and what a stronger approach looks like, based on what you actually did.

  • Every action, in order

    What was checked, who was asked, what was changed and when. Nothing is overwritten.

  • Scores by skill

    Each exercise scores the skills it was built to test, from what was actually done.

  • Weak habits, flagged

    Poor patterns in how the work was done are flagged, not only wrong final answers.

  • A replay of the attempt

    Step through the session as it happened, with typed personal details masked.

Attempt recordIllustration
  1. You're paged
  2. Investigate
  3. Mitigate
  4. Communicate and review
4 steps recordedReplay available

Questions

Frequently asked questions

Anything else? Read the general questions.

What is a site reliability engineer?

An engineer who keeps software services running: monitoring them, responding when they break, and making them more reliable over time.

Do I need to know a specific cloud provider?

No. The simulated company has its own services and tools, so the skills carry over to any stack.

Ready to try it?

Create an account and start your first exercise. It runs in your browser.

Start practising