Feature 02 · Follow-ups

An AI interviewer that asks the second question

Mentari is an AI mock interview with follow-up questions: instead of reading out a list, the interviewer listens to each answer and responds to it. Say you improved something and it asks by how much. Offer a solution before clarifying the constraint and it notices. It interrupts when you drift, gives a single hint when you are genuinely stuck, and does not keep drilling a candidate who is struggling — the judgement a good human interviewer uses.

Why the follow-up is the interview

Anyone can prepare an answer to "tell me about a project". The follow-up — why that approach, what breaks at ten times the load, what you personally did, how you measured the result — cannot be memorised, because it depends on what you just said. It is where interviewers find out whether the first answer was real.

Practice that never pushes back trains the wrong skill: delivering a prepared answer to a listener who accepts it.

What you say, and what comes next

If your answer…A question list does thisMentari does this
Claims a result with no numberNext questionAsks how much, from what baseline, measured how
Says "we" throughoutNext questionAsks what you personally did
Jumps to a solutionNext questionAsks what you assumed about the constraints
Drifts off the questionWaitsInterrupts and brings you back
Stalls completelyWaitsOffers one hint, then moves on if it does not help
Is clear and specificNext questionRaises the difficulty

How Mentari decides when to push

The round is planned in advance as a set of topics with a time budget. Within each topic the interviewer probes until it has the evidence it was looking for — a quantified result, the reasoning behind a choice, your individual contribution — or until the topic has had its share of the clock, and then it moves on.

If you say you have not worked on something, it switches to a different thread rather than scoring you on a topic you never claimed. If you go quiet, it nudges. If you are stuck, it offers one hint and sees what you do with it.

Voice and text

By voice it is a two-way conversation: it talks back and can interrupt. By text it is a chat with an interviewer who follows up on what you write. Both use the same plan and produce the same scorecard.

The other two things that make it different

Last updated · Written by the Mentari team

Questions

Things people actually ask.

01Is it harsh?

No. It is persistent, not hostile. It pushes on vague answers and backs off when a candidate is genuinely stuck, because drilling someone a third time teaches nothing.

02Can I ask it to go easier or harder?

The difficulty adapts to how you are answering, and the round is set at the level of the role you named.

03How is this different from a chatbot I prompt myself?

A chatbot you prompt is an interview you are running. Here the interviewer holds the plan and the clock, and you do not know what is coming — which is most of what makes a real interview hard.

Find out before the interviewer does.

Your first interview is free. No card needed. Upload your résumé, name the role and company, and you are in the room in about a minute.