AI mock interview

Data analyst mock interview

A data analyst mock interview on Mentari is a 20- or 40-minute voice or text interview built from your résumé, the job description and research on the company you are applying to. The AI interviewer asks how you would define and investigate a metric, has you reason through SQL and analysis choices aloud, questions the numbers on your résumé, and returns a scorecard where each of six competency scores is tied to a quote from your answers.

Who this is for

Students and freshers targeting analyst roles, and working professionals moving into analytics from engineering, operations or finance.

What a data analyst interview tests — and how Mentari runs each part

RoundWhat it testsWhat Mentari does
Résumé and projectsWhether your analysis projects had a question, a method and a resultAsks what decision your analysis changed, and how you know
SQL and data reasoningJoins, aggregation, window logic — and knowing when a result is wrongHas you talk through a query approach, then introduces duplicates or nulls
Metrics and product senseDefining a metric, and diagnosing why it movedGives you a metric that dropped and follows your investigation step by step
StatisticsSampling, significance, correlation versus causation in plain wordsAsks you to explain a result to a manager who does not know statistics
BehaviouralCommunicating findings, handling pushback from stakeholdersPlays the sceptical stakeholder

You do not pick these from a menu. Mentari plans the round from your résumé, the job description and what it finds about how the company interviews for the role, so the mix changes with every session.

Typical questions — and the follow-up that comes next

Lists of interview questions are everywhere. What decides the interview is the second question: the one that checks whether the first answer was real. These are typical of the role, each paired with the follow-up a good interviewer asks.

  1. 01Daily active users dropped 8% this week. How do you investigate?

    Follow-up · You have found it is only on Android. What is the next cut, and what would make you stop digging?

  2. 02How would you find duplicate customer records in a table?

    Follow-up · Two rows differ only in the case of the email address. Are they duplicates, and who decides?

  3. 03What is the difference between a LEFT JOIN and an INNER JOIN?

    Follow-up · Tell me about a time the wrong join silently changed a number you reported.

  4. 04Walk me through an analysis on your résumé.

    Follow-up · What decision was made differently because of it?

  5. 05How would you measure whether a new feature is successful?

    Follow-up · The metric went up but revenue did not. What do you tell the product manager?

  6. 06Explain a p-value to someone who is not technical.

    Follow-up · They now ask, "so is it true or not?" What do you say?

  7. 07Which chart would you use to show this, and why?

    Follow-up · What would a misleading version of the same chart look like?

  8. 08Your dashboard and the finance team’s number disagree. What do you do?

    Follow-up · It turns out yours is the wrong one. How do you handle that?

Where candidates lose marks

  • Jumping to a query before asking what the business question is.
  • Listing tools (Excel, SQL, Power BI, Python) instead of describing a result.
  • Treating every correlation on a dashboard as a cause.
  • Giving an investigation plan with no stopping point.

What your scorecard shows

About thirty seconds after you finish, you get a score out of 10 and six competency scores. Each one is tied to a verbatim quote from your transcript, so you can see the exact answer that earned or cost the marks. How scoring works.

CompetencyWhat it measures
CommunicationWhether your answers were clear, direct and easy to follow — not your accent.
Technical depthHow well you understand the substance of the role, beyond definitions.
Problem solvingHow you break down an unfamiliar problem and reason towards an answer.
OwnershipWhat you personally did, decided and were accountable for.
StructureWhether your answers had a shape: context, action, result, and a point.
Company fitWhether you showed the signals this particular company looks for in this role.

Practise for another role

Last updated · Written by the Mentari team

Questions

Things people actually ask.

01Does the AI check my SQL syntax?

No. It is a conversational interview, so it evaluates how you reason about data — which tables, which joins, which edge cases, how you would verify the result. Practise syntax on a query sandbox; practise explaining your thinking here.

02Is it useful if I am switching into analytics?

Yes. It reads your résumé, so it will ask how your current experience transfers, which is exactly the question a hiring manager has for a career switcher.

03What do I get afterwards?

A score out of 10, six competency scores each tied to a quote from your transcript, the risks a real interviewer would flag, and a drill list for the next attempt.

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.