Every candidate gets a real machine of their own.

Not a code box in a browser tab. An isolated cloud environment, created for one session, with the tools the job actually runs on.

Hugo, a backend candidate, in his own environment

Live environment

Running
  • vmlab-vm-01Ready
  • svcapi-serverReady
  • lbgatewayProvisioning

2h 12m left of the 3h limit

lab terminal

$ docker compose up -d

3 services started

$ systemctl status api

active (running)

$ curl -s localhost:8080/health

{"status":"ok"}

Improvised test environments were the problem

Without Scalyz

A take-home that runs on the candidate's laptop, with a different setup every time.
A test environment somebody has to build, host, then clean up.

With Scalyz

The same machine, the same tools, the same starting point for everyone.
Created on its own when the session starts, deleted on its own when it ends.

A live environment, created for one session

It starts when the candidate starts. Real infrastructure, real services, and it belongs to that one candidate until the session is over.

Inside the environment

Ready
  • Terminal

    docker · systemctl · curl

  • Services

    api-server, gateway

  • Files and editor

    The project, ready to open

Everything the job runs on, already installed

  • A machine of their own

    One isolated cloud environment per candidate, running from the first minute of the session.

  • The tools of the job

    A terminal, running services, files to open. Nothing simulated, nothing stubbed out.

  • A clock that runs itself

    The time box opens when they start and closes on its own. No scheduling, nobody looking over their shoulder.

The environment lives for the session and nothing longer. When the clock runs out, or the candidate hands the work in, the machine and everything on it is destroyed.

Everyone starts from the same place

Every candidate in a campaign gets the same environment and the same time limit. Same image, same services, same starting state. Identical conditions are what makes two scores worth comparing.

A head of talent

Backend hiring, July

3h time box
  • Camille Durandlab-vm-01
  • Hugo Moreaulab-vm-01
  • Léa Fontainelab-vm-01

Same image, same tools, same clock, for all three

Campaign results

  1. 1Camille DurandRecommended84
  2. 2Lucas Bernard77
  3. 3Léa Fontaine71
An engineering team whose own systems stay untouched

Environment lifecycle

  1. 14:02Environment created
  2. 14:02Session started, clock running
  3. 17:02Environment deleted, automatically

Nothing left running, nothing to clean up

Isolated from your systems, and gone afterwards

The environment runs on our infrastructure, never on yours. Candidates never reach your network, your repositories, or your data. When the session closes, the machine is torn down for you.

  • Nothing to provision, nothing to clean up
  • Candidates never touch your systems
  • Every environment is destroyed when its session closes

A day at work, not an exam hall

Candidates work on their own hours, inside the deadline you set. They open a terminal, read the task, and get on with it, the way they would on a Tuesday morning.

A candidate working in her own environment, on her own schedule
Her own hoursThe tools she knowsOne clock, shared by everyone

Give every candidate a machine of their own.

Run your first assessment this week. The environments create and delete themselves.

One credit per candidate, no seats to buy, credits never expire