For a long time, engineering interviews had this unspoken rule:
Close ChatGPT.
Turn off Copilot.
How dare you use AI to code! That's not real engineering.
We're not doing that.
If you're interviewing for an engineering role at Kogan, you can use AI. Cursor, Claude, ChatGPT, Copilot, whatever's already part of your workflow. Honestly, we'd rather see how you actually work with these tools.
Why we changed it
Our engineers use AI constantly. It helps write, debug, refactor, and unblock work across Kogan.com, Mighty Ape and the rest of the group.
So banning it for sixty minutes during an interview never really made sense.
The old style of interview mostly measured memorisation. The ability to recall patterns you've seen before and reproduce them under pressure.
That's not the job anymore.
What matters now is judgement. How you think. How you work with the tools. How you spot when something looks off. How you make decisions when the model gives you three different answers and one of them is subtly wrong.
That's a much more useful signal for us.
What this means in practice
Use the tools you'd normally use at work.
Paste the problem in.
Generate code.
Debug with AI.
Refactor with it.
We'll ask you to share your screen and talk through your thinking as you go, similar to how you'd walk another engineer through a PR or problem you're solving together.
Using AI won't count against you.
Blindly trusting it probably will.
What we're actually looking for
We're not measuring how quickly you can type out a perfect solution from memory.
We're looking at:
- how you approach the problem
- how you use AI as a tool instead of a crutch
- whether you can tell good suggestions from bad ones
- how you debug and validate output
- whether you understand the code you're shipping
- how you handle the moment the model confidently gives you nonsense
The engineers who get the most leverage from AI aren't the ones pasting prompts into ChatGPT all day. They're the ones applying judgement on top of it.
That's the skill.
A few simple rules
Use whatever language or stack you're strongest in. We care far more about how you think than whether you memorised framework trivia.
Talk through your reasoning. Silent copy paste sessions don't tell us much.
If you're stuck, say it. Good engineers ask good questions.
And don't ship code you can't explain. If we ask why something works, "the AI wrote it" isn't really an answer.
The reality
We run lean teams and move fast. AI tools help our engineers operate with more leverage than ever before, and the people who do well here tend to be the ones who adapt quickly and use the tools well.
So if you're sharper with AI than without it, that's completely fine.
Honestly, we'd expect you to be.