In these candidate reports, PayPal's Data Scientist interviews open with a practical, SQL-heavy coding screen. Later rounds vary between product metrics and A/B testing cases, take-home modeling reviews, ML system design, and project deep dives.
$146K
Avg. Base Comp
$214K
Avg. Total Comp
Not reported
Typical Rounds
Not reported
Process Length
PayPal's Data Scientist interview leans hard on practical SQL. In these reports, the first technical round was often a live HackerRank screen. One candidate spent about 45 minutes on SQL alone, working through window functions, advanced joins, and aggregations. Their questions included customer counts by country split into merchant and consumer customers, and percentile buckets for transaction value that had to include customers with zero transactions. They also faced cleanup problems involving dates stored as strings, a null customer ID, and leading spaces. Another candidate got three medium SQL problems heavy on subqueries and window functions. They also got pandas, numpy, and general Python prompts, such as cleaning a student_id column before computing a weighted average. Interviewers tolerated small syntax slips but cared about logical correctness, and the pace felt time-pressured.
Later rounds varied. One candidate had a product metrics case and an A/B testing round focused on experiment design. They also had a hiring-manager conversation about a favorite project and working with a non-technical stakeholder. Another completed a take-home prediction task on tabular data and then defended their modeling and feature choices, including how to approach classification without labels in the training data. That was followed by a manager round built around a business problem. The candidate who received an offer went through three rounds:
Prepare by writing window-function and subquery SQL quickly on messy data, practicing common pandas cleaning and aggregation, and rehearsing clear explanations of why you made each choice in your own projects. Outcomes were mixed, and two candidates noted slow or absent HR follow-up after their rounds.
Synthesized from 13 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 Paypal process.
Overall, the conversations were more collaborative than I expected. I felt most confident during the technical and case study rounds because they closely matched the kind of credit risk analytics and client-facing work I do every day. The biggest surprise was how much the interviewers cared about my thought process rather than arriving at the "perfect" answer—they often dug into why I made certain assumptions or how I would handle ambiguity. The behavioral interview was where I started sweating a bit. Instead of standard questions, they challenged me with situations around influencing stakeholders without authority, handling disagreements, and learning from mistakes. It felt less like reciting prepared stories and more like having an honest conversation, which forced me to think on my feet. Looking back, that was probably the most valuable part of the entire process.
Questions asked: build a model for detecting credit risk
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 Paypal
Write a query to get the total three-day rolling average for deposits by day
| Question | |
|---|---|
| Merge Sorted Lists | |
| String Shift | |
| Paired Products | |
| Over-Budget Projects | |
| Swipe Precision | |
| Find the Missing Number | |
| Third Purchase | |
| Bank Fraud Model | |
| Total Spent on Products | |
| Bagging vs Boosting | |
| P-value to a Layman | |
| Variable Error | |
| Nearest Common Ancestor | |
| String Mapping | |
| Sort Strings | |
| Precision and Recall | |
| Finding The Mode | |
| Target Indices | |
| Poker Pair | |
| Priority Queue Using Linked List | |
| Success Measurement | |
| Assumptions of Linear Regression | |
| Duplicate Rows | |
| Testing Price Increase | |
| Nightly Job | |
| Descending Alphanumeric Sorting | |
| Unsafe Content ML Design | |
| Most Repetition | |
| Concurrent LLM Serving |
Synthesized from candidate reports. Individual experiences may vary.
Two candidates mention a recruiter call or shorter phone screen alongside their technical rounds. One candidate's phone and behavioral conversations included standard questions such as why they were interested in the role.
A SQL-heavy technical screen appears in most reports, often run live on HackerRank; one candidate's lasted about 45 minutes. Candidates describe window functions, subqueries, joins, and messy data cleanup tasks, sometimes paired with Python, pandas, or numpy questions. Another candidate's coding round covered SQL, Python, and statistics. The candidate who received an offer had one easy-to-moderate Python question and one SQL question.
One candidate described a product metrics case and a separate A/B testing round focused on experiment design. Another had a manager interview built around a business problem, talking through assumptions, tradeoffs, and how to frame the analysis for stakeholders.
A modeling-focused stage appears in several reports, though the format varies. One candidate received a take-home prediction task on tabular data and later defended their modeling and feature choices. The candidate who received an offer completed an end-to-end ML system design discussion spanning data preparation through deployment, followed by a deep dive into past projects. Interviewers pushed on why a specific approach was chosen over alternatives.
Hiring-manager and behavioral conversations appear in several loops. Reported questions include a favorite project, a time working with a non-technical stakeholder, and why you are interested in the role. One candidate also had a longer technical interview centered on explaining the projects they were currently handling.
