Descartes Underwriting Data Scientist one candidate reports an HR call, a week-long notebook take-home, a discussion and math brainteasers, then a team on-site.
$125K
Avg. Base Comp
$150K
Avg. Total Comp
4 rounds
Typical Rounds
Not reported
Process Length
For a Descartes Underwriting Data Scientist interview, one candidate reported a four-stage sequence that began with an HR call and then moved to a take-home exercise containing three notebooks. The final notebook required building a binary classification model, and the candidate had one week to complete the assignment. Prepare to explain your modeling work, not just submit it: the subsequent interview focused largely on the final notebook, alongside personal-experience questions and math-based brainteasers.
That makes it worthwhile to rehearse a concise walkthrough of a binary-classification project: the choices you made, what the model output means, and how you would discuss your approach in conversation. Be ready to connect prior experience to the work rather than treating the technical exercise as isolated. The reported process ended with an on-site meeting involving most team members, so practice explaining the same work clearly to people with different perspectives. Descartes’ underwriting data-science work is connected to risk modeling, but the reported interview did not specify domain questions. This guide reflects one intern-level report, so details may vary by level and location.
Synthesized from 1 candidate report by our editorial team.
Had an interview recently?
Share your experience. Unlock the full guide.
Real interview reports from people who went through the Descartes Underwriting 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 Descartes Underwriting
Select the 2nd highest salary in the engineering department
| Question | |
|---|---|
| Top Three Salaries | |
| Experiment Validity | |
| Delivery Estimate Model | |
| Instagram TV Success | |
| Group Success | |
| P-value to a Layman | |
| Four Person Elevator | |
| Random Forest Explanation | |
| Causal Email Journey | |
| Three Zebras | |
| Recruiting Leads | |
| Xgboost vs Random Forest | |
| Spam Classifier | |
| Green Dot | |
| Facebook Job Board Design | |
| Data Stream Median | |
| Facebook Story Success | |
| Your Strengths and Weaknesses | |
| A/B Test Power Size | |
| Bounds for Coffee and Tea Drinkers | |
| Delivery Online | |
| Building Lyft Line | |
| Empty Neighborhoods | |
| Rolling Bank Transactions | |
| Customer Orders | |
| Comments Histogram | |
| Closest SAT Scores | |
| Employee Salaries | |
| Subscription Overlap |
Synthesized from candidate reports. Individual experiences may vary.
One candidate reported an HR call as the first stage. The report does not specify questions or duration, so prepare a clear account of your background and interest without assuming a particular screening format.
The candidate received a take-home test with three notebooks and reported having one week to finish it. The last notebook involved building a binary classification model, making a reproducible and explainable approach especially useful.
Candidates report an interview covering personal experience, discussion of the take-home, and brainteasers. The brainteasers were described as mostly math-based, while discussion concentrated mainly on the final notebook.
The reported final stage was an on-site interview with most team members. The report gives no further agenda, but you may need to communicate your take-home reasoning consistently across several conversations.
