Travelers data scientist candidates describe a recruiter screen, conversational technical rounds on modeling and statistics fundamentals, and behavioral questions. Reports range from a two-round internship process to longer loops, so the format varies by role and level.
$138K
Avg. Base Comp
$160K
Avg. Total Comp
Not reported
Typical Rounds
Not reported
Process Length
Candidates describe the Travelers data scientist interview as a test of how clearly you explain your thinking. One candidate who accepted an offer said the technical round was discussion-based with no live coding. It covered missing values, median versus mean, and how XGBoost differs from a random tree model. Another candidate was also asked to compare XGBoost and Random Forest, in a one-hour round that included Python coding.
Open-ended reasoning questions come up repeatedly. A full-time candidate was asked how to count traffic lights in a city's downtown. An internship candidate was asked a similar Fermi question about New York City and was also asked to define p-values and Type I versus Type II errors. A senior-leaning candidate got a geometric probability problem about a rope cut into three pieces forming a triangle. Say your assumptions out loud in all of these.
Process length varies a lot by role. Reports range from a two-round internship process to much longer multi-round loops, and a manager-level candidate describes four 45-minute sessions, most of them behavioral. Expect to walk through projects end to end and explain your own contribution. Be ready for "why Travelers" and career-goal questions. One candidate who did not get an offer was asked how the insurance industry will change, and said their answer was too general. Tie your view to concrete data problems.
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 Travelers process.
I went through a recruiter screen first, then a technical interview with two data scientists that lasted about an hour. The recruiter and DSLP manager conversation was mostly behavioral, and the final round mixed behavior, coding, and data science discussion. The technical interview opened with a pretty standard behavioral question about whether I prefer routine tasks or non-routine work, then moved into Python coding and broader data science topics. I also got asked to compare XGBoost versus Random Forest, so they did want to see that I could talk through model tradeoffs and not just code. In the same round, there was a consulting-style section that felt a little odd to me, with questions that seemed more about how I think through ambiguous business problems than about pure technical depth.
Prep tip from this candidate
Be ready for a one-hour round that combines Python coding, SQL, and model comparison questions like XGBoost versus Random Forest. It also helps to practice explaining your thinking on ambiguous consulting-style prompts, since that part seemed to matter more than I expected.
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 Travelers
How would you assess the validity of the result?
| Question | |
|---|---|
| Bagging vs Boosting | |
| Xgboost vs Random Forest | |
| Hurdles In Data Projects | |
| Type I and II Errors | |
| Client Solution Pushback | |
| Kindergarten Feasibility | |
| Your Strengths and Weaknesses | |
| Delivery Fees | |
| 2nd Highest Salary | |
| Employee Salaries | |
| Top Three Salaries | |
| P-value to a Layman | |
| Random SQL Sample | |
| Booking Regression | |
| Rectangle Overlap | |
| Size of Joins | |
| Always Excited Users | |
| WAU vs Open Rates | |
| Total Spent on Products | |
| Delivery Estimate Model | |
| Instagram TV Success | |
| Group Success | |
| Covariance vs Correlation | |
| Assumptions of Linear Regression | |
| Lasso vs Ridge | |
| Get Top N Frequent Words | |
| Distribution of 2X - Y | |
| Fair Coin | |
| Integer String Addition |
Synthesized from candidate reports. Individual experiences may vary.
Candidates report starting with a recruiter conversation of roughly 20 to 30 minutes. It typically covers availability, a resume walkthrough and one project. One candidate described it as a short chat, and one said the recruiter stayed responsive afterward. Have a concise account of your own contribution ready.
Several candidates describe an early conversation with the hiring manager or a DSLP manager. It is mostly about past work and how it fits the role, and it can be largely behavioral. Candidates report being asked to explain projects end to end and what they specifically contributed, so prepare a clear and honest account.
Candidates describe technical rounds of about an hour with two data scientists or managers. One reports no live coding and a discussion of missing values, median versus mean and XGBoost. Another reports Python coding plus an XGBoost versus Random Forest comparison. An internship candidate reports p-values and Type I and II errors. Format may differ by role and level.
Candidates report open-ended questions such as counting traffic lights downtown, a geometric probability problem about a rope forming a triangle, and a consulting-style business prompt. They appear to test how you structure ambiguous problems and state assumptions more than whether you reach an exact answer, so narrate your reasoning.
Candidates report behavioral segments covering why Travelers, why the role, career goals, teamwork and culture fit. One candidate describes a four-part loop of programming, statistics, case study and behavioral rounds. A manager-level candidate describes four 45-minute sessions and says those interview rounds wrapped up within two weeks, with offer discussions taking longer.
