Novartis data scientist interviews, as candidates describe them, combine recruiter and manager conversations, panels, and sometimes a written exam, group discussion, or final presentation. Reports are mixed and all four ended without an offer.
$138K
Avg. Base Comp
$160K
Avg. Total Comp
2-6 rounds
Typical Rounds
Not reported
Process Length
Reports of the Novartis data scientist interview vary, so ask your recruiter what format to expect. Four reports cover the role or a close variant, and all four ended without an offer. They do not describe one fixed loop.
One candidate reports a four-round sequence: a 40-question technical exam covering statistics, machine learning, deep learning and quantitative methods, then a group discussion in groups of eight on whether AI is a boon or bane. A resume-focused technical interview and an HR behavioral round followed. That candidate says breadth mattered more than depth in one area.
Another candidate describes a recruiter screen, a hiring manager conversation, a senior team member chat, two panels on one day, and a final presentation. That candidate reports a posting focused on machine learning, real-world evidence and predictive modeling, and tied the presentation to the job description. The same candidate noticed the recruiter described the role differently from the posting, so confirm the actual scope early.
A third candidate reports phone screens, a full on-site day and an HR call. The questions included defining a project plan and explaining how to build one, and the candidate was told time had run out for questions at the end. A fourth, who interviewed for an expert-level role, reports an interviewer who cut in repeatedly during a string-array comparison question. Practice stating assumptions in a sentence before coding.
Prepare broad ML and statistics fundamentals, a clear walkthrough of your own projects, and a short list of questions you can ask quickly.
Synthesized from 6 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 Novartis process.
I went through a four-round technical and interpersonal gauntlet and didn't advance past it, but the structure was clear enough that I got a sense of where I fell short.
First came a technical exam: forty questions spanning statistics, machine learning, deep learning, and quantitative methods. Forty questions is a lot under time pressure, and the breadth requirement meant that depth in one area wasn't enough if you're weak elsewhere. After that was a group discussion where they split us into groups of eight and gave us a topic to debate—something like "Is AI a boon or bane?" It was different from typical technical interviews; they seemed to be watching how you communicate and handle disagreement in a group, not just how you think.
Round three was a technical interview focused entirely on your resume—specific projects and methods you've actually used. That one felt more grounded than the exam, and I was more confident walking out of it, but by that point the first two rounds had probably already shaped how I was perceived.
The final round was standard HR behavioral questions. Nothing surprising there, just the formality at the end of a long process.
The whole thing tested you across multiple dimensions—breadth of technical knowledge, communication and collaboration, real project experience, cultural fit. I think they weight all of that together, and no single round is make-or-break. That said, the technical exam is a high bar if you haven't maintained fundamentals across the full ML stack.
Prep tip from this candidate
Expect a 40-question technical exam testing breadth across stats, ML, deep learning, and quant—depth in one area won't cut it. Study fundamentals broadly before interviewing, and be ready for a group discussion on a topic like 'Is AI a boon or bane,' where they're assessing communication and collaboration skills.
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 Novartis
What do you tell an interviewer when they ask you what your strengths and weaknesses are?
| Question | |
|---|---|
| 2nd Highest Salary | |
| Monthly Customer Report | |
| Cumulative Distribution | |
| Experiment Validity | |
| Last Transaction | |
| Weighted Keys | |
| Hurdles In Data Projects | |
| Brain Cancer Treatment Outcomes | |
| Always Excited Users | |
| Retailer Data Warehouse | |
| Total Spent on Products | |
| P-value to a Layman | |
| RMS Error | |
| Reducing Error Margin | |
| Detecting ECG Tachycardia Runs | |
| Size of Joins | |
| Fair Coin | |
| Cumulative Reset | |
| Causal Email Journey | |
| Time Difference | |
| Greatest Common Denominator | |
| Random Forest Explanation | |
| Subscription Retention | |
| Secret Wins | |
| Sum to Zero | |
| Missing Housing Data | |
| Valid Anagram | |
| Xgboost vs Random Forest | |
| Licensing Valuation |
Synthesized from candidate reports. Individual experiences may vary.
Two candidates report starting with a recruiter or phone screen. One says the recruiter's description of the role differed from the posting, which asked for machine learning, real-world evidence and predictive modeling. Confirm the scope and the later stages with the recruiter, since the reports are not uniform.
One candidate describes an informal, mission-focused hiring manager conversation followed by a senior team member who knew their work. Another describes a full day with the team that included a project-planning question. You may be asked to explain past work and how you plan projects, so prepare examples from your own experience.
Formats reported vary. One candidate describes a 40-question exam covering statistics, machine learning, deep learning and quantitative topics. Another reports two panels on one day, one more technical and open-ended, which gave room to explain analytical methods. Review fundamentals broadly and be ready to walk through the methods you have used.
Late-stage steps differ by report. One candidate describes a group discussion in groups of eight, then a resume-based technical interview and an HR behavioral round. Another describes a final presentation and then waiting for HR to confirm a decision. You may need to follow up on timing, so ask about next steps at the end.
