Meesho Data Scientist candidates report a broad process spanning SQL, DSA coding, statistics, machine learning depth, PySpark, and detailed discussion of resume projects.
$130K
Avg. Base Comp
$146K
Avg. Total Comp
2 rounds
Typical Rounds
Not reported
Process Length
Meesho Data Scientist interviews reported here test a deliberately broad mix: SQL and coding sit alongside statistics, machine-learning fundamentals, and close discussion of projects. One candidate described an online assessment before two interviews; their assessment included statistics, SQL, and one PySpark question. The first interview combined a resume walkthrough, two DSA questions, and broad ML, while the second went deeper on projects and behavioural fit.
Prepare for a technical switch between SQL, algorithms, and ML reasoning. Reported SQL prompts include top-three salaries by department, median calculations with CTEs or window functions, rolling sums, and partitioning. Coding examples range from logistic regression implemented from scratch and Asteroid Collision to Combination Sum II and a graph problem suited to union-find. Be ready to explain your approach, complexity, and assumptions rather than treating the task as a library-only exercise.
ML discussion can be mathematical as well as practical. Candidates reported p-value versus power, error types, precision and recall, ROC-AUC, regularization, model trade-offs, and model equations connected to the resume. Project conversations may probe feature engineering, loss functions, debugging choices, imbalanced data, leakage, monitoring, A/B testing, and concept drift. Practice explaining the technical choices in your own work clearly to both technical and non-technical listeners.
The evidence is strongest for early technical interviews, so later-stage structure is not consistently established.
Synthesized from 5 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 Meesho process.
The first round ran for about 90 minutes and moved across coding, SQL, maths and statistics, resume projects, ML/DL fundamentals, and PySpark. The DSA question was an equation-satisfaction graph problem where union-find was the right approach.
The ML discussion was mathematical: I had to explain p-value versus statistical power, Type I versus Type II errors, and false positives versus false negatives. I was also asked how to move a tabular problem from XGBoost to a deep-learning setup, including cross-entropy and regularization, and to discuss bias in recommender systems. Questions went into equations and derivations for models on my resume. On the data side, they covered RDDs, DataFrames, lazy execution, UDFs, and left anti and semi joins. SQL required clear solution structure for medium-to-advanced problems.
I did not clear the technical stage and did not receive an offer.
Prep tip from this candidate
Prepare to derive the equations behind every model on your resume, and review p-value versus power, error types, deep-learning approaches to tabular data, and recommender-system bias. Also drill medium-to-advanced SQL solution structure, union-find and DP problems, plus Spark concepts including lazy execution, UDFs, RDDs, and anti/semi joins.
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 Meesho
Select the 2nd highest salary in the engineering department
| Question | |
|---|---|
| Top Three Salaries | |
| Compute Deviation | |
| Average Order Value | |
| Top 3 Users | |
| Scrambled Tickets | |
| Size of Joins | |
| Sequentially Fill in Integers | |
| Identifying User Sessions | |
| Precision and Recall | |
| Hurdles In Data Projects | |
| Cumulative Sales By Product | |
| Target Indices | |
| Equal Binary Subarrays | |
| Hidden Culprit | |
| Filling Supermarket Bag | |
| Ride-Sharing App Schema | |
| NxN Grid Traversal | |
| Upsell Carousel | |
| Shortest Path Algorithms | |
| Logistic Regression from Scratch | |
| Duplicate Product Names | |
| Median Household Income | |
| Evaluating Revenue Decline | |
| Empty Neighborhoods | |
| Rolling Bank Transactions | |
| Customer Orders | |
| Comments Histogram | |
| Closest SAT Scores | |
| Experiment Validity |
Synthesized from candidate reports. Individual experiences may vary.
One candidate reported an assessment before interviews covering statistics and SQL, with one PySpark question. Their SQL examples included chi-square goodness of fit, median calculations using a CTE or window function, partitioning, cumulative rows, and rolling sums.
Candidates report technical conversations that can mix a resume walkthrough with DSA coding, SQL, statistics, ML/DL fundamentals, and PySpark. Reported tasks include logistic regression from scratch, top-three salaries by department, Asteroid Collision, Combination Sum II, and a union-find graph problem.
One candidate who completed two interviews reported a later discussion focused on ML depth, resume projects, and behavioural fit. Candidates may be asked to explain model internals, loss functions, feature engineering, debugging choices, trade-offs, and false-positive versus false-negative decisions.
