Oracle Data Scientist interviews, as three candidates describe them, combine online assessments or HackerRank tests, Python and algorithm coding, ML fundamentals, scenario-based hypothesis questions, and project discussion.
$126K
Avg. Base Comp
$184K
Avg. Total Comp
Not reported
Typical Rounds
Not reported
Process Length
Oracle's Data Scientist interview does not look uniform. Three candidates describe three different setups, so prepare for range rather than one fixed script.
One candidate describes a four-round process: an HR screen, an hour-long technical screen with a principal applied data scientist about past projects, and then three back-to-back interviews with applied and principal scientists. In that loop, one interview centered on Python and a LeetCode-style stock buying and selling problem. Another was scenario-based, asking how to approach hypothesis testing on problems the team actually works on. The third covered team fit. That candidate also got a data warehouse architecture question and felt the technical focus was scattered.
A second candidate began with an online assessment covering ML concepts, SQL, probability and two Python problems. Later rounds, about 60 minutes each, included a behavioral conversation, a case on designing a data analysis pipeline for trading data, a coding round, a code review exercise and a presentation of their own ML projects. Topics included the Central Limit Theorem, checking normality, and imbalanced data.
A third candidate describes a single HackerRank assessment: a data structures and algorithms problem, two optional ML notebooks, and multiple-choice questions on data science concepts and logic, including number sequences and code time complexity. A sample test is available beforehand.
All three reported no offer, so use these accounts as preparation signals, not as a scoring guide. Practice medium-level Python algorithms, ML fundamentals such as overfitting and imbalanced data, and basic statistics and SQL. Also rehearse explaining the trade-offs in your own projects and how you would generate and test hypotheses on a real problem.
Synthesized from 3 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 Oracle process.
The whole process ran on HackerRank, and for me it was a single online assessment. It opened with a DSA problem, and there were two optional machine learning notebooks for anyone interested in the AI side of the role. The rest was a set of multiple-choice questions covering data science concepts and general logic. The logic questions stood out: I was shown a sequence of numbers and had to work out the next value, and I was also asked about the complexity of a given piece of code.
Before the real assessment, candidates can take a sample test on HackerRank that covers data structures, data cleaning, and algorithms. Treat it as a preview of what the actual assessment looks like.
I didn't receive an offer after that assessment. For prep, I'd spend time on the sample test and on quick reasoning for number sequences and time complexity, since those came up alongside the coding problem.
Prep tip from this candidate
Drill time complexity of short code snippets and number-sequence pattern questions, and work through the HackerRank sample test covering data structures, data cleaning, and algorithms before the real assessment.
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 Oracle
Select the 2nd highest salary in the engineering department
| Question | |
|---|---|
| Closest SAT Scores | |
| Subscription Overlap | |
| Upsell Transactions | |
| Merge Sorted Lists | |
| String Shift | |
| Largest Salary by Department | |
| Minimum Change | |
| Scrambled Tickets | |
| Rectangle Overlap | |
| Project Pairs | |
| Integer to Roman | |
| Find the First Non-Repeating Character in a String | |
| Bagging vs Boosting | |
| Size of Joins | |
| The Brackets Problem | |
| Nearest Common Ancestor | |
| Normalize Grades | |
| Basic Regex | |
| Random Forest Explanation | |
| Resumable Fact Table Load | |
| Cyclic Detection | |
| Sort Strings | |
| Binary Tree Conversion | |
| Valid Anagram | |
| Find Duplicate Numbers in a List | |
| Hurdles In Data Projects | |
| Finding The Mode | |
| Production Model Monitoring | |
| Dijkstra implementation |
Synthesized from candidate reports. Individual experiences may vary.
Candidates report different openings. One had an HR screening, while another took an online assessment with ML, SQL and probability multiple-choice questions plus two Python problems. A third described a HackerRank test with a DSA problem, optional ML notebooks, and logic questions. The entry point may vary.
One candidate reports an hour-long technical screen with a principal applied data scientist. It focused on past projects, experience and relevance to the role. Another candidate presented their own machine learning projects, discussing challenges such as overfitting and imbalanced data. Be ready to explain trade-offs precisely.
Two candidates report coding questions. One recalled a medium-difficulty stock buying and selling problem under time pressure, and another described LeetCode-style algorithmic problems that were not extremely difficult. The HackerRank candidate also reported a DSA problem and time-complexity questions.
One candidate reports scenario-based questions about hypothesis testing on real team problems. Another candidate got a trading-data pipeline design case, a code review exercise, and theory topics such as the Central Limit Theorem, normality and imbalanced datasets. Expect applied reasoning, not only definitions.
Two candidates describe behavioral or fit-oriented conversations. One covered background and motivation, and the other included a team fit and cultural alignment interview in a loop of three back-to-back sessions. Outcomes reported across all three accounts were no offer.
