Ebay Data Scientist candidates report SQL and Python coding, with machine learning discussion, project deep dives and case-style questions in some rounds. Reported round counts range from two to three.
$157K
Avg. Base Comp
$200K
Avg. Total Comp
2-3 rounds
Typical Rounds
Not reported
Process Length
Across the four Data Scientist reports, the technical core is SQL and Python. One candidate started with a CodeSignal round of Python and SQL questions and was then asked to attend the remaining interviews in person; the questions included a Python groupby problem and a gaps-and-islands problem about finding streaks. Another describes a first round on pandas and SQL with a lot of groupby and aggregation work. A third got a SQL coding problem and a more algorithmic Python round, and a fourth discussed a SQL window-function question.
Machine learning shows up in different forms. One candidate was asked about resume projects, including DBSCAN, LDA and Gen AI, and why they chose DBSCAN over k-means. Another was given a design question about predicting user classification, where the conversation centered on feature engineering, model choice and trade-offs rather than code. One report also mentions a case study and some basic math problems.
The format varies. One candidate counts three rounds: Python and SQL, then a project discussion with SQL and a case-style segment, then a project deep dive. The first was virtual and the others were in the office. Another counts two rounds that felt more syntax-focused than the prep email suggested, with production-style pipeline scenarios layered on top. A third had an HR screen followed by 45-minute technical rounds with the data science team, plus behavioral questions such as why you want to join eBay.
For prep, practice groupby, aggregations, window functions and streak problems, and rehearse a structured machine learning walkthrough from features to evaluation. Be ready to defend every method on your resume and to explain your reasoning aloud. Two of the four reports ended in no offer and two candidates were still waiting to hear back.
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 Ebay process.
The process started with an HR screening call to go over my background and interest in the role. They asked about my work experience and gave me a heads-up that the technical rounds would follow a predictable format — each round was 45 minutes with members of the data science team. After that initial conversation, HR reached out to schedule and prep me for what was coming, which felt organized at least.
The technical interviews were split across SQL and Python. In one round, I got a SQL coding problem to work through — nothing too exotic, but enough to see how I'd approach a real schema problem. The Python round was more algorithmic in nature. There was also a machine learning systems design question that came up, where they asked me how I'd approach predicting user classification — less about writing code and more about explaining your thinking on feature engineering, model choice, and trade-offs. They were curious about my technical depth too, asking me to rate my SQL and Python skills on a scale and whether I used PyTorch frequently.
Beyond the technical stuff, there were behavioral questions: Why do you want to join eBay? What's the soonest you could start? Pretty standard. The whole interview felt efficient but honestly quite by-the-book. No real surprises or moments where I felt like they were digging deep into my actual project experience. Looking back, I wish I'd spent more time really thinking through a concrete machine learning problem end-to-end — the systems design piece wasn't a full technical interview, more of a discussion, and I could have shown more rigor there. I didn't get an offer, and while I can't say for certain why, I suspect my approach to the ML design question wasn't as structured as they wanted.
Prep tip from this candidate
Focus on solid SQL problem-solving and be ready to articulate a machine learning system design from feature engineering through model selection and evaluation — that's more of a discussion than coding, so practice explaining your reasoning clearly rather than just writing queries.
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 Ebay
Write a query to get the total three-day rolling average for deposits by day
| Question | |
|---|---|
| Merge Sorted Lists | |
| Experiment Validity | |
| Button AB Test | |
| Longest Streak Users | |
| Rectangle Overlap | |
| Find the First Non-Repeating Character in a String | |
| Random Bucketing | |
| P-value to a Layman | |
| Google Maps Improvement | |
| Nearest Common Ancestor | |
| Hurdles In Data Projects | |
| Target Indices | |
| Testing Price Increase | |
| Matrix Rotation | |
| New UI Effect | |
| Sample Size Bias | |
| Listing Bookings Aggregation | |
| Max Width | |
| Bias vs. Variance Tradeoff | |
| Inherited Model Evaluation | |
| String Palindromes | |
| Walking Robot | |
| Impossibly Iterative Fibonacci | |
| The Longest Journey | |
| Seller Type Modeling | |
| Azure Kubernetes Infrastructure | |
| Relational Migration | |
| Free Shipping Mention Test | |
| Processing Large CSV |
Synthesized from candidate reports. Individual experiences may vary.
One candidate reports an HR call that covered their work experience and interest in the role, followed by scheduling and a heads-up that the technical rounds were 45 minutes each with members of the data science team. Behavioral questions in that account included why you want to join eBay and how soon you could start. Other reports do not mention a screen.
One candidate reports a CodeSignal round of Python and SQL questions before the in-person interviews. Reported questions include a pandas groupby problem, a gaps-and-islands problem about finding streaks, and a SQL coding problem on a schema. One candidate describes the Python round as more algorithmic, and another discussed a SQL window-function question.
Candidates report being asked about resume projects, including DBSCAN versus k-means, LDA and Gen AI. One was asked how they would predict user classification, covering feature engineering, model choice and trade-offs. Another describes a case study, basic math problems and a project deep dive. One account adds production-style pipeline scenarios.
One candidate counts three rounds, with the first virtual and the rest in the office; another counts two rounds. Two candidates report no offer and two were still waiting at the time of their report. The reports do not give an end-to-end timeline, so confirm the schedule and format with your recruiter.
