Based on three candidate reports, EPAM Systems' Data Scientist interviews open with a recruiter/HR screening; one Senior Data Scientist account adds a theoretical ML exam and an NLP-focused technical round, two candidates reported scheduling breakdowns, and none reported an offer.
$128K
Avg. Base Comp
$180K
Avg. Total Comp
5 rounds
Typical Rounds
Not reported
Process Length
Three candidate reports for EPAM Systems' Data Scientist roles point to a few recurring themes: screening calls that focus on background and past projects, technical rounds that didn't always match what the candidate was told to prepare for, and inconsistent follow-through from recruiters once the initial conversation ends.
Two candidates specifically flagged reliability problems between the screening call and the next stage. One applicant, interviewing for a Data Scientist role out of the KL, Malaysia office, was told an online assessment invite would arrive by end of day, then heard nothing for three working days; when they followed up, the recruiter said requirements had shifted to senior-level candidates with 5+ years of experience, even though the original posting stayed live. A second candidate, targeting a Data Scientist role on the Odysseus team, had their interview rescheduled twice, then had the interviewer never join the confirmed call at all. Neither candidate reached a technical round, since the process broke down before that stage.
A Senior Data Scientist candidate did complete a full loop and offers the clearest technical picture available. After a supportive HR screening, they hit a theoretical machine learning exam that didn't match the site's description of a data-science-and-system-design assessment. The next stage, a 90-minute technical interview, was billed as system design but turned into a deep dive on NLP topics — transformers, LSTMs, and a question on implementing positional encoding — with limited discussion of the candidate's own projects. A behavioral round on handling conflicting stakeholder requests followed, and the loop closed with an interview with the Head of AI. Notably, post-interview feedback referenced production deployment and monitoring topics that never came up in any of the actual conversations, leaving the candidate unsure what had actually been evaluated.
Given this gap between what interviewers say to expect and what's actually asked, candidates should prepare broadly across core ML fundamentals and neural network architecture rather than trusting the stated format alone, and should confirm scheduling details more than once before an interview date.
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 Epam Systems process.
I had scheduled an interview for a Data Scientist role on the Odysseus team, but the experience was deeply frustrating. The interview was rescheduled twice before we even got to a confirmed time. When I finally locked in a date and time with the recruiter, I cleared my morning and waited for the call — but no one showed up. The interviewer didn't connect, and when I tried to reach the recruiter for clarity, they were unavailable. I was left hanging with no explanation and no way to reschedule. It felt unprofessional and disrespectful of my time.
From what I could gather from others who did make it through their interviews, the process involved a two-person technical panel covering live coding in R and SQL, along with domain-specific questions around healthcare data models like OHDSI and OMOP — particularly defining cohorts and working with their common data model. After the technical round, there was a conversation with the department head about experience and role fit. On the technical side, candidates reported being asked SQL and Python questions, with an emphasis on machine learning depth and some high-level deep learning concepts. The expectation was working code for even simple problems, so preparation needed to be solid. Given how the process fell apart for me, I never got to that stage, but I'd recommend anyone interviewing there to follow up proactively with the recruiter well before your scheduled time to confirm everything is still on track.
Prep tip from this candidate
If you make it past scheduling, expect live SQL and Python coding questions with a focus on healthcare data models (OHDSI/OMOP cohort definition) — but confirm your interview details multiple times with the recruiter before the day of, as scheduling reliability appears to be an issue.
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 Epam Systems
Create top_ads with the top 3 ads and return the row counts for inner, left, right, and cross joins with ads
| Question | |
|---|---|
| Level Of Rain Water In 2D Terrain | |
| Classification and Regression | |
| SageMaker Deployment Architecture | |
| Weighted Average Sales | |
| Best DAU | |
| 2nd Highest Salary | |
| Empty Neighborhoods | |
| Top Three Salaries | |
| Merge Sorted Lists | |
| Employee Salaries | |
| Rolling Bank Transactions | |
| Customer Orders | |
| Comments Histogram | |
| Closest SAT Scores | |
| Subscription Overlap | |
| First to Six | |
| Monthly Customer Report | |
| Upsell Transactions | |
| Experiment Validity | |
| Download Facts | |
| Cumulative Distribution | |
| First Touch Attribution | |
| Prime to N | |
| Manager Team Sizes | |
| Top 3 Users | |
| Random SQL Sample | |
| Compute Deviation | |
| Find the Missing Number | |
| Largest Salary by Department |
Synthesized from candidate reports. Individual experiences may vary.
The process typically opens with a recruiter or HR conversation. One Data Scientist candidate described a straightforward walkthrough of their background and past projects, with little technical depth at this stage. A Senior Data Scientist candidate similarly described the HR round as supportive, saying it set a positive tone for the rest of the process. Expect this stage to focus on fit and experience rather than technical evaluation.
Two candidates reported breakdowns between the initial screening and the next stage. One Data Scientist candidate was told an online assessment invite would arrive by end of day, then received no communication for three working days; when they followed up, the recruiter said the role's requirements had shifted to senior-level candidates, even though the same posting remained live. Another candidate, targeting a Data Scientist role on the Odysseus team, had their interview rescheduled twice, then had the interviewer never join the confirmed call. Confirm scheduling details directly with your recruiter more than once before your interview date.
A Senior Data Scientist candidate described a technical test that didn't match the website's description of a data-science-and-system-design assessment. Instead, it was a heavily theoretical traditional machine learning exam. This account comes from a single Senior Data Scientist report, so scope and difficulty may vary by role and interviewer, but it's a signal not to assume the posted format will match what's actually tested.
In the same Senior Data Scientist process, the interviewer said upfront to expect system-design questions, but the 90-minute session instead focused heavily on NLP topics, including transformers and LSTMs. One question asked the candidate to explain how they'd implement positional encoding in an LSTM. The candidate said there was only a glancing look at their own projects and real-life ML cases. This is one candidate's account of a single interview, so treat it as an example of the depth expected rather than a guaranteed topic list.
The Senior Data Scientist process included a behavioral round asking how the candidate would manage conflicting stakeholder requests. This is based on a single candidate's report, so treat it as one example of the kind of scenario question that can come up rather than a fixed script, but it's worth preparing a concrete example of navigating competing priorities.
The Senior Data Scientist candidate's process closed with an interview with the Head of AI, described as the final stage before a decision. Post-interview feedback referenced production deployment and model monitoring, topics that hadn't come up in any of the actual interview conversations, which left the candidate uncertain about what was being evaluated. This mismatch is based on one candidate's account, so use it as a reminder to ask directly what feedback is based on if it doesn't match your own experience.
