Candidate reports on the Axa Data Analyst interview describe a conversational intro, a CV-driven SQL and Power BI discussion, and, in one case, a walk-in drive with background and reinsurance questions.
$74K
Avg. Base Comp
$85K
Avg. Total Comp
2 rounds
Typical Rounds
Not reported
Process Length
The Axa Data Analyst interview, as candidates describe it, rewards people who know their own work well. One candidate who received an offer reports two rounds. The first was a behavioral conversation about their background and why they were interested in insurance. The second was technical, and the interviewer drew questions directly from the CV. SQL and Power BI Desktop were the focus, and the candidate describes the questions as grounded in real analytics work rather than abstract algorithm puzzles. Preparation should therefore start with your own projects. Be ready to explain the queries you wrote, the visualization problems you solved in Power BI, and why you made the technical choices you did.
A second candidate, who did not receive an offer, describes a different format: a walk-in drive in Gurgaon where candidates waited about two hours and were then interviewed in small groups. For a Data Analyst opening, the conversation stayed mostly on background. Questions covered why they left previous roles, why they wanted AXA, and a few reinsurance questions. This candidate was told that shortlisted people would receive an assessment link within 24 hours, but heard nothing for weeks. If you attend a walk-in, confirm how and when you will be contacted.
The two accounts point to the same preparation. Have a clear story about your career moves and your interest in AXA and insurance. Refresh SQL and Power BI fundamentals. Learn the basics of the insurance business, including reinsurance. The formats differ by hiring route, so check which one applies to you.
Synthesized from 3 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 Axa process.
The process had three main rounds. The first was a fairly standard conversation with HR about my background, motivation, and why I wanted to work at AXA Climate. The second was the most demanding: a technical and business-focused interview where I had to explain how I would approach a data problem, challenge the assumptions, and communicate the result to non-technical stakeholders. The final round was with the hiring manager and members of the team, and it felt more like a real working discussion than an interview.
What surprised me was that they were not only testing whether I could analyse data. They wanted to see whether I could connect the analysis to the business and explain it clearly. I felt confident discussing my previous projects and walking through my reasoning. I started sweating when they pushed back on some of my assumptions and asked what I would do if the data were incomplete or contradicted the initial hypothesis. There was no perfect answer, so I had to think out loud and stay structured.
The atmosphere was professional but genuinely friendly. By the final round, I stopped trying to give polished “interview answers” and focused on showing how I actually work: asking questions, being honest about uncertainty, and turning messy information into something useful.
Questions asked: From what I remember, the technical questions were practical rather than textbook-style. They asked me to describe a data project from start to finish: how I clarified the business question, cleaned the data, chose the right metrics, validated the result, and presented the conclusion. They also challenged me on what I would do if the dataset contained missing values, duplicates, outliers, or conflicting information from different sources.
On SQL, the questions covered joins, aggregations, filtering, and how I would structure a query to calculate a metric by customer, product, or time period. There were follow-up questions about checking whether a join had accidentally duplicated rows and how I would investigate a result that looked unexpectedly high or low. For Python, the discussion was mainly around using pandas to clean, transform, and analyse data rather than writing complicated algorithms.
The case prompt was broadly about receiving an imperfect dataset and turning it into a useful business recommendation. I had to explain which questions I would ask first, how I would explore the data, what assumptions I was making, and which charts or KPIs I would use. They kept adding complications—missing fields, limited historical data, and stakeholders wanting an answer quickly—to see how I would prioritise.
They also asked questions such as:
The take-home instructions were quite open-ended, which was probably the hardest part. The goal was not simply to produce as many charts as possible. I needed to identify the important story in the data, document my assumptions, and present a short, actionable recommendation. The lack of detailed instructions initially felt strange, but I realised that ambiguity was part of the test. They wanted to see how I framed the problem and communicated my choices, not just whether I could write code.
One small detail that mattered: during the presentation, they asked “why?” repeatedly. Why this metric? Why this visualisation? Why exclude those records? The analysis itself was important, but being able to defend each decision was just as important.
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 Axa
In which case would you use a bagging algorithm versus a boosting algorithm
| Question | |
|---|---|
| Stakeholder Communication | |
| Your Strengths and Weaknesses | |
| 2nd Highest Salary | |
| Employee Salaries | |
| Random SQL Sample | |
| Booking Regression | |
| P-value to a Layman | |
| Size of Joins | |
| Hurdles In Data Projects | |
| Random Forest Explanation | |
| Lasso vs Ridge | |
| Type-ahead Search | |
| Scalped Ticket | |
| WAU vs Open Rates | |
| Three Zebras | |
| Missing Housing Data | |
| Assumptions of Linear Regression | |
| Xgboost vs Random Forest | |
| Success Measurement | |
| Target Indices | |
| RAG Strict Source Control | |
| Data Preparation for Imbalanced Data | |
| Classification and Regression | |
| HHT or HTT | |
| Duplicate Rows | |
| Netflix Churn Prediction | |
| Modifying a Billion Rows | |
| Type I and II Errors | |
| Second Ace |
Synthesized from candidate reports. Individual experiences may vary.
One candidate who received an offer reports a conversational first round about their background and why they were interested in the insurance sector. It was described as a standard introduction to assess fit. Candidates should be ready to explain their path and motivation for joining Axa.
The same candidate reports that the second round drew questions directly from their CV, with a heavy focus on SQL and Power BI Desktop. They walked through SQL queries and visualization problems, and found the round felt like a real conversation about skills. Expect to defend your own project choices in detail.
A different candidate describes a walk-in drive where HR had candidates wait about two hours and then called them in groups of two or three. Questions centered on previous roles, reasons for leaving each, why AXA, and a few reinsurance questions. This format may apply only to certain hiring drives, so confirm the process with the recruiter.
The first candidate reports receiving an offer after the two rounds and says the process moved fairly quickly. The walk-in candidate was told an assessment link would come within 24 hours for shortlisted candidates, but reports no message at all after four weeks. Follow-up experiences may vary by team and hiring route, so ask for a timeline and a contact.
