Two candidates describe Samsung ML Engineer interviews that mix a coding question, ML fundamentals such as bias and variance, and a deep dive into past projects. Both report receiving offers.
$203K
Avg. Base Comp
$248K
Avg. Total Comp
3 rounds
Typical Rounds
2-3 weeks
Process Length
Both candidates who describe the Samsung Electronics ML Engineer interview report a mix of algorithm coding, machine learning theory and detailed discussion of past work. Both also report receiving offers.
One candidate describes a recruiter call after a referral, then a 1-hour hiring manager phone screen that opened with a LeetCode-style question. A 40-minute panel with three interviewers followed, centered on presenting past projects. The other candidate describes three rounds: an introductory conversation, a 30-minute medium-level graph problem followed by ML questions, and a final round combining a project deep dive with more advanced algorithm questions.
Bias and variance came up in both accounts, so be ready to explain it clearly. One candidate was asked about regularization and model complexity, and about diagnosing why a model performs poorly on training data. Other reported topics were batch normalization, gradient descent and a detailed explanation of how LSTM works.
Both accounts also include a deep dive into past projects, so prepare to defend your work. Be ready to explain why you chose one model over another, what trade-offs you weighed and what you would change in hindsight. One candidate was also asked about culture fit and commute distance. For coding, practice medium-level graph and data-structure problems.
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 Samsung Electronics process.
I went through a pretty straightforward but technical process for the Machine Learning Engineer role. It started with a recruiter call after a referral, and that went smoothly enough that they moved quickly into the interviews. The first round was a 1-hour phone screen with the hiring manager, and it opened with a LeetCode-style coding question before shifting into my background, technical skills, and relevant experience. That part felt like a mix of general software problem solving and role-specific discussion, since they also wanted to hear about computer science fundamentals and how I approach data structures and algorithms.
The later round was more of a panel format with three interviewers and The candidate, lasting about 40 minutes. I had to present past projects and then answer questions about them, which made it feel more like a deep dive than a standard behavioral chat. They asked about my major, culture fit, and attitude, and even asked where I lived to gauge commute distance to the workplace. On the technical side, the questions were very focused on machine learning and deep learning, including a detailed explanation of how LSTM works and why I chose one deep learning model over another. I also got a basic theory question on bias and variance, so they were checking both fundamentals and whether I could justify modeling choices in context. Overall, the process was fair but definitely not light on ML depth. I ended up getting the offer, and my main takeaway was to be ready to explain past projects clearly and defend model selection decisions, not just recite theory.
Prep tip from this candidate
Be ready to explain LSTM in depth and to justify why you chose one deep learning model over another for a past project. It also helps to prepare a concise project presentation, since they spent a full round asking Q&A on prior work and fit.
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 Samsung Electronics
Why can the same machine learning algorithm produce different success rates on the same dataset
| Question | |
|---|---|
| Valid Anagram | |
| Hurdles In Data Projects | |
| Bias - Variance Tradeoff and Class Imbalance in Finance | |
| Cyclic Detection | |
| Target Value Search | |
| Bias vs. Variance Tradeoff | |
| Food Delivery Times | |
| Data Preparation for Imbalanced Data | |
| String Palindromes | |
| Deciding Between Solutions | |
| Shortest Path Algorithms | |
| Text Editor With OOP | |
| Seller Type Modeling | |
| Impossibly Iterative Fibonacci | |
| Client Solution Pushback | |
| Your Strengths and Weaknesses | |
| LRU Cache 1 | |
| Regularization and Validation | |
| Gradient Descent Calculation | |
| Bias Variance Tradeoff | |
| Merge Sorted Lists | |
| Find the Missing Number | |
| Maximum Profit | |
| Scrambled Tickets | |
| The Brackets Problem | |
| Prime to N | |
| Bagging vs Boosting | |
| One Element Removed | |
| Random Forest Explanation |
Synthesized from candidate reports. Individual experiences may vary.
Candidates report starting with a recruiter call, in one case after a referral, or a casual introductory conversation about background and projects. One candidate says the recruiter signaled that later rounds would become more technical. Be ready to summarize your projects and ML experience briefly.
One candidate reports a 1-hour hiring manager phone screen that opened with a LeetCode-style question before moving to background and fundamentals. Another describes a roughly 30-minute medium-level graph problem. Practice timed data structure and algorithm questions, since both accounts include them.
Candidates report foundational ML questions in the technical conversations. Reported topics include bias and variance, regularization, why a model might underperform on training data, batch normalization and gradient descent. Prepare to explain each concept and how you would apply it to a real model.
Both candidates describe a late round centered on past projects. One reports a roughly 40-minute panel with three interviewers, covering an LSTM explanation, model selection and culture fit. The other reports a final round that added harder algorithm questions. Prepare a concise project walkthrough and be ready to justify your decisions.
