An AI-powered interview simulator that lets students rehearse out loud, get honest feedback in the moment, and walk into the real room already knowing what their answers sound like.
The problem
Final-year students spend months polishing a résumé and about ten minutes preparing to actually speak. Then the call comes, the question lands, and the answer that felt fine in their head falls apart out loud.
The options available to them are all broken in the same way. Rehearsing in front of a mirror gives no feedback. Friends and family are too kind to be useful. The campus career centre does run mock interviews — with a two-week waiting list and one session per student per term.
In our discovery interviews the same sentence kept coming back: "I know what I want to say, I just can't say it under pressure." The gap wasn't knowledge. It was reps.
How might we give students unlimited, honest interview practice they can do alone at 11pm — and still walk away knowing what to fix?
My approach
As product owner I held the backlog — which meant every phase ended with something cut, not just something added.
{{ ph.body }}
Collaboration
We ran two-week sprints with a demo at the end of each one. My job was to make sure the two developers never had to guess what "done" meant.
{{ rit.body }}
Critical thinking
Every one of these started as a disagreement inside the team. Being product owner meant I had to argue the user's side with evidence, not opinion.
The solution
One loop, repeatable at any hour: answer, hear how it landed, try again better.
The outcome
{{ res.stat }}
{{ res.label }}
{{ res.body }}
"I'd already answered that question four times at home. So when he asked it, I wasn't thinking about what to say — I was just saying it."
— test participant, two weeks after a successful graduate interview
Reflection & takeaways
{{ out.stat }}
{{ out.body }}
Next case study