Razorpay data analyst candidates report an online assessment, SQL-heavy technical rounds, and RCA or product case studies. Here is how the reported rounds differed and what to practice.
$25K
Avg. Base Comp
$34K
Avg. Total Comp
2-4 rounds
Typical Rounds
Not reported
Process Length
Across five candidate reports, SQL is the one constant in Razorpay's analytics interviews. Candidates describe medium-difficulty problems over tables, plus questions on joins, window functions, running or cumulative sums, self joins and, in one case, SQL indexing and order of execution. One candidate says the SQL tested whether they understood the logic rather than the syntax.
The second layer is business reasoning. Several candidates describe root cause analysis or product case studies, such as diagnosing why a key metric dropped or why a hotel business had declining bookings. One candidate says they jumped to solutions before laying out hypotheses and lost ground. Another says the case discussion mattered more than a puzzle they answered correctly. State your hypotheses and how you would check them before proposing fixes.
The rest varies by candidate. One reports an online assessment mixing aptitude, probability, combinatorics and SQL. Another reports two DSA problems plus SQL in an hour. Basic Python, A/B testing, Tableau, resume questions and company knowledge each came up for at least one candidate. One candidate was asked to design a dashboard around an AI agentic workflow.
Outcomes in these reports were mostly rejections or silence: one offer, three no-offers, and one process put on hold. Several candidates describe poor communication, including HR going quiet and rescheduled slots. Confirm each interview time in writing, and don't read silence as a verdict on your performance.
Synthesized from 5 candidate reports by our editorial team.
Had an interview recently?
Share your experience. Unlock the full guide.
Real interview reports from people who went through the Razorpay process.
The question that stuck with me was to build a dashboard using an AI agentic flow. That's a different kind of prompt from the usual BI exercise, and it meant thinking about how an agent would gather data, make decisions, and feed results into a visual output, not just how to lay out charts. I had prepared for standard dashboarding and SQL, so this one took some adjusting.
The process itself was hard to follow. After agreeing on dates and times with the recruiter, I kept receiving invitations with different schedules, and corrections came slowly. This happened across several rounds. Some interviews were postponed shortly before they were due to start, usually citing panel unavailability or an interviewer on sick leave. A few interviewers joined very late, and some didn't join at all. Some rounds seemed to have no clear purpose, and several questions didn't connect to the role.
I cleared all the earlier technical rounds, but my final technical round was rescheduled three times before it was cancelled. I was then told the hiring process had been put on hold for all positions. I didn't get an offer. The lack of communication was the most frustrating part of the whole process.
Prep tip from this candidate
Be ready to design a dashboard around an AI agentic workflow, covering how the agent gathers, decides, and presents data. Also confirm every interview slot in writing with the recruiter, since invitations here frequently conflicted.
Share your own interview experience to unlock all reports, or subscribe for full access.
Sourced from candidate reports and verified by our team.
Topics based on recent interview experiences.
Featured question at Razorpay
In which case would you use a bagging algorithm versus a boosting algorithm
| Question | |
|---|---|
| Hurdles In Data Projects | |
| Z and t-Tests | |
| Production Rollout Challenges | |
| Reddit-like Notifications | |
| Empty Neighborhoods | |
| Rolling Bank Transactions | |
| 2nd Highest Salary | |
| Top Three Salaries | |
| Customer Orders | |
| Comments Histogram | |
| Closest SAT Scores | |
| Experiment Validity | |
| First to Six | |
| Monthly Customer Report | |
| Download Facts | |
| First Touch Attribution | |
| Last Transaction | |
| Random SQL Sample | |
| Compute Deviation | |
| Top 3 Users | |
| Button AB Test | |
| Employee Salaries (ETL Error) | |
| Minimum Change | |
| Lowest Paid | |
| 500 Cards | |
| Raining in Seattle | |
| Bank Fraud Model | |
| Subscription Overlap | |
| Month Over Month |
Synthesized from candidate reports. Individual experiences may vary.
Candidates report an online first screen, but its format differs. One describes a 90-mark Hackerearth test with aptitude, probability, combinatorics and two SQL questions. Another describes a HackerRank test with SQL, probability and aptitude. A third describes two array-based DSA problems plus SQL in an hour. Prepare SQL plus light quantitative practice.
Candidates report a live technical round that usually opens with a short discussion of past work, then moves to SQL. Reported topics include joins, window functions, cumulative sums, self joins, indexing and SQL execution order. Interviewers were described as supportive, nudging candidates when they got stuck, so talk through your logic out loud.
Several candidates report a case-based round or section, which may include a metric-drop question, a declining-bookings scenario or a puzzle. One candidate who answered the puzzle correctly felt the case reasoning mattered more. A structured approach is typically safer: clarify the metric, list hypotheses, say what data you would check, then recommend actions.
Candidates report mixed additional topics: basic Python, A/B testing, Tableau visualization, resume deep-dives, company knowledge and strengths-and-weaknesses questions. One candidate was asked how to present the same analysis to a Product Manager versus leadership. These topics did not appear in every report, so treat them as likely follow-ups rather than guaranteed rounds.
Candidates report uneven communication after interviews. Two say HR went silent after their rounds. One candidate reports conflicting invitations, rescheduled or cancelled rounds, and then a hiring hold across positions. One candidate received an offer. Confirm timings in writing and follow up with the recruiter if you hear nothing.
