
Two candidates describe S&P Global data science interviews that combined recruiter and behavioral conversations with technical discussion of NLP, machine learning, feature engineering and past projects. Experiences differed on structure and outcome.
$125K
Avg. Base Comp
$160K
Avg. Total Comp
3 rounds
Typical Rounds
Not reported
Process Length
The two candidate reports we have for S&P Global data science roles point in the same direction: be ready to explain your own work in depth. One intern candidate had a recruiter call, a behavioral round with the hiring manager built around their resume and projects, and a technical interview that centered on NLP. That candidate was asked to walk through an NLP method, covering its use case, strengths and weaknesses, and how the model works. They described it as a conceptual conversation rather than a whiteboard coding exercise.
A senior candidate described a different experience. They went through three rounds. The first was behavioral, covering teamwork and remote work, with a manager, team members and a director. The second covered coding fundamentals, feature engineering, machine learning and deep learning concepts, and a deep dive into past projects. They found the third round repeated much of the second.
The practical takeaway is to prepare to defend your projects and one or two methods in depth, including why you chose them and where they fail. The senior candidate also said the recruiter's guidance on format was vague and contradictory, so ask for a round-by-round outline and confirm it with the hiring manager. Outcomes differed: the intern received an offer after a long wait, while the senior candidate reported no feedback after the final round. With only two reports, treat the process as varying by team and level.
Synthesized from 2 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 S&P Global process.
I started with a recruiter call that felt pretty straightforward — they wanted to get a sense of my background and make sure I was a fit for the role. After that passed, I had a behavioral round with the hiring manager where we mostly walked through my resume and projects. Nothing too intense, but it gave them a chance to understand what I'd actually built and whether I could articulate my thinking clearly.
The technical interview came next, and since I'd mentioned NLP work in my background, that's where they focused. They asked me to walk through a method I'd learned in NLP — explain the use case, the strengths and weaknesses, how the model actually works. It wasn't a coding challenge on a whiteboard; it was more about demonstrating real understanding of the concepts and being able to defend my choices. The conversation felt natural, and the interviewer seemed genuinely interested in my reasoning rather than just looking for a "right answer."
The whole process took about three weeks from my first call to final interview, which felt reasonable. The one hiccup was that after my interviews wrapped up, the recruiter went quiet for two months before getting back to me with an offer. That wait was frustrating, but the offer did come through. The experience felt collaborative overall — they were testing whether I could think through problems and communicate clearly, not just whether I could code under pressure.
Prep tip from this candidate
Be ready to deeply explain an NLP method you've actually worked with or studied — focus on use cases, model mechanics, strengths, and weaknesses. This role tests conceptual understanding and communication more than rapid coding, so practice articulating your reasoning clearly.
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 S&P Global
Find the names of all employees who joined before their manager
| Question | |
|---|---|
| LRU Cache 1 | |
| Hurdles In Data Projects | |
| Credit Score Estimation | |
| User Journey Analysis | |
| RAG Strict Source Control | |
| Testing Constraints | |
| Concurrent LLM Serving | |
| Interpolating Missing Temperatures | |
| Why Do You Want to Work With Us | |
| Your Strengths and Weaknesses | |
| 2nd Highest Salary | |
| Empty Neighborhoods | |
| Rolling Bank Transactions | |
| Employee Salaries | |
| Closest SAT Scores | |
| Merge Sorted Lists | |
| Subscription Overlap | |
| Comments Histogram | |
| Top Three Salaries | |
| Find the Missing Number | |
| Last Transaction | |
| Cumulative Distribution | |
| Monthly Customer Report | |
| Size of Joins | |
| Payments Received | |
| Top 5 Turnover Risk | |
| Experiment Validity | |
| Button AB Test | |
| Paired Products |
Synthesized from candidate reports. Individual experiences may vary.
Both candidates report starting with a recruiter. The intern candidate describes a straightforward conversation about background and fit. The senior candidate says the recruiter gave vague and conflicting information about format and topics, so asking for a round-by-round breakdown may help.
The intern candidate describes a behavioral round with the hiring manager that walked through their resume and projects. The senior candidate reports behavioral questions on teamwork and remote work from a manager, team members and a director, with heavily overlapping questions. Expect to explain your work clearly.
The intern candidate was asked to explain an NLP method they had learned, including use case, strengths, weaknesses and model mechanics, in a conceptual discussion rather than a coding challenge. The senior candidate reports coding fundamentals, feature engineering, machine learning, deep learning concepts and project deep dives.
Outcomes differed. The intern candidate says the process took about three weeks from first call to final interview, then waited about two months for an offer. The senior candidate reports a third round that repeated much of the second, followed by no feedback. Timing after interviews may vary.