MSD data scientist interviews reported by candidates combine a recruiter call, project discussion, and an on-site or in-person session. Expect to explain your reasoning aloud, and for some roles to handle domain biology questions.
$135K
Avg. Base Comp
$150K
Avg. Total Comp
2-3 rounds
Typical Rounds
Not reported
Process Length
Candidates describe MSD data science interviews as conversational rather than adversarial, but the content varies a lot by team. Two candidates who reported on data scientist-type roles both ended without an offer, so treat this as a small sample.
One candidate for a general Data Scientist role started with a recruiter call, then gave an online presentation on past work. After about a month, they were invited to an office session. They completed a take-home beforehand, presented their solution in person, and then did live coding. They described both problems as easy to average, with the emphasis on explaining your thinking rather than raw difficulty. The session lasted about an hour and a half.
A second candidate, for a Senior Data Scientist and Computational Biologist role, reported a recruiter screen covering motivation, salary expectations, and Python. It also included questions on protein structure prediction. The on-site with the hiring manager was a discussion of protein modeling, how protein expression is measured, and how to bring wet lab data into predictive models. Model validation and Python visualization libraries also came up. This candidate found the biology the hardest part and waited more than two and a half weeks for a generic rejection.
If you are applying to a biology-adjacent team, ask the recruiter how the role splits between modeling and domain science. Whatever the team, prepare to walk through your own projects in detail, defend modeling choices out loud, and be comfortable with Python and SQL basics. An intern who received an offer also described project-focused technical rounds.
Synthesized from 4 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 Msd process.
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 Msd
Select the 2nd highest salary in the engineering department
| Question | |
|---|---|
| Monthly Customer Report | |
| Cumulative Distribution | |
| Experiment Validity | |
| Last Transaction | |
| Weighted Keys | |
| Always Excited Users | |
| Brain Cancer Treatment Outcomes | |
| Retailer Data Warehouse | |
| Hurdles In Data Projects | |
| 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 | |
| 85% vs 82% |
Synthesized from candidate reports. Individual experiences may vary.
Candidates report a recruiter call early on. It covered standard HR topics, motivation for joining MSD, and in one case salary expectations and Python experience. One candidate for a biology-oriented role also got domain questions on protein structure prediction. Expect it to be relatively low pressure.
One Data Scientist candidate reports an online interview built around a presentation of previous work, with engaged follow-up questions about what they did and why. Teams may differ, but preparing a clear, concise walkthrough of a past project, including your decisions and results, is a sensible way to get ready.
One candidate completed a take-home assignment after waiting about a month for the next invitation. They described the problem as easy to average in difficulty. This step was reported by only one candidate, so it may not apply to every team, but be ready to defend your approach.
The in-person session for one candidate had two parts: presenting the take-home solution, then live coding, over roughly an hour and a half. Another candidate's on-site with a hiring manager was a discussion of protein modeling, expression measurement, wet lab data in models, and ML validation. Interviewers focused on reasoning.
Both data scientist candidates reported no offer. One received a generic rejection after more than two and a half weeks, with no reply to a request for feedback. Timelines may vary, so plan for a possibly slow and quiet follow-up.
