Oscar Insurance Data Scientist one candidate reports a behavioral conversation, an open-ended morbidity-classification take-home, and a lengthy final session covering the submission, pragmatic technical work, and team conversations.
$146K
Avg. Base Comp
$189K
Avg. Total Comp
3 rounds
Typical Rounds
Not reported
Process Length
The reported Oscar Insurance Data Scientist process moved through three distinct stages: a behavioral first conversation, an open-ended take-home, and a final session described as a six-hour gauntlet. The take-home asked the candidate to create features for a morbidity-classification problem using supplied data, with roughly six hours available. Prepare to explain how you framed an ambiguous classification task, selected and evaluated features, and made tradeoffs under a time limit.
In the final session, the candidate reviewed the take-home and completed custom technical questions built around member-routing optimization. The questions reportedly built on one another and emphasized pragmatic NumPy functions rather than elaborate data-structures-and-algorithms exercises. Practice communicating your assumptions as a problem evolves, keeping code readable, and tying each decision back to the optimization objective. Oscar describes data work across areas including disease modeling and risk management, which makes the health-insurance framing relevant to this report. One candidate report is available, so team-specific details may differ.
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 Oscar Insurance 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 Oscar Insurance
Explain what a p-value is to someone who is not technical
| Question | |
|---|---|
| Confidence Interval Explanation | |
| Risk Assessment Model | |
| Why Do You Want to Work With Us | |
| 2nd Highest Salary | |
| Employee Salaries | |
| Experiment Validity | |
| Prime to N | |
| Top 5 Turnover Risk | |
| Bagging vs Boosting | |
| Size of Joins | |
| Rectangle Overlap | |
| Always Excited Users | |
| Hurdles In Data Projects | |
| Total Spent on Products | |
| Booking Regression | |
| The Brackets Problem | |
| Distribution of 2X - Y | |
| Google Maps Improvement | |
| Fair Coin | |
| Covariance vs Correlation | |
| Valid Anagram | |
| Integer String Addition | |
| Type-ahead Search | |
| Assumptions of Linear Regression | |
| Sort Strings | |
| String Mapping | |
| Three Zebras | |
| Lasso vs Ridge | |
| Precision and Recall |
Synthesized from candidate reports. Individual experiences may vary.
One candidate reports an easy behavioral first round. Prepare concise examples of collaboration, judgment, and how you communicate analytical work; no further behavioral prompts or evaluation criteria were reported.
The candidate reports an open-ended take-home using provided data to create features for a morbidity-assessment classification model, with an approximately six-hour limit. Be ready to explain feature choices, modeling approach, and the tradeoffs made within that constraint.
The reported final stage lasted six hours and included a take-home review, technical assessment, Socratic dialogue, personality assessments, and conversations across multiple teams. Technical questions reportedly built on a member-routing optimization problem and used pragmatic NumPy-style functions rather than elaborate DSA problems.
