Meesho Data Scientist-1 Interview Experience | Bangalore
- Location
- Bangalore
- Posted
- 15 May 2025
Summary
I participated in the Meesho Data Challenge 2024, won, and was invited to interview for a Data Scientist-1 role in Bangalore. The interview process consisted of four rounds covering DSA, SQL, Statistics, ML/DL, system design, and project discussions. I received a confirmation from HR on the same day as my final interview round.
How it went
Participated in Meesho Data Challenge 2024 (won!) and hence got the chance to interview for open roles at Meesho.
HR contacted on 10th April 2025 and shared the interview details:
- Round 1: DSA, SQL, Statistics & Mathematics, Basic ML/DL/GenAI Knowledge
- Round 2: ML Depth
- Round 3: ML Breadth (case study based)
- Round 4: Overall Discussion
Round 1 (17/04/2025)
- What do you understand by p-value?
- If we fail to reject the null hypothesis then do we accept an alternate hypothesis?
- We want to sample values between [0, 1] uniformly but it should remain within the unit circle. How do you do?
- Explain bias-variance tradeoff
- Training on a large dataset, training loss is getting reduced but validation loss is stagnant, how will you address it?
- Coding
-
Given a list of citizens in a country, Info of birth date and death date, Find the max population throughout the history
-
Find the max number of libraries that can be installed from ‘libaries_needed’: You are given three lists: Libraries_needed = [‘pandas’, ‘numpy’] Libraries_wd_no_preq = [‘A’, ‘B’, ‘C’] Libraries_wd_preq = [[‘A’, ‘B’], [‘B’]]
Here 0th index elements are prerequisites for ‘pandas’ and so on
-
- Explain ReLU activation function
- What is the drawback of using ReLU
- Explain positional encoding in transformers
- Why did authors use sine and cosine? Why can’t we use binary values?
Round 2
-
First (21/04/2025)
- Explain layer and batch normalization
- Discussed potential alternate solution of Meesho Hackathon project
- Convex and non-convex loss functions
- CNNs (projects are CV based)
- Discussion on pooling techniques
- Segmentation basics and pooling techniques in that
- Classification metrics: class imbalance, comparison between ROC-AUC and PR-AUC
- If loss is NaN then possible reasons?
-
Second (25/04/2025) : This was an additional interview with another panel.
- Bias-variance tradeoff
- Regularization techniques
- Why does L1 regularization create sparsity?
- Optimal batch size and why?
- Why do we go into the negative of the gradient?
- Maximum Likelihood Estimation and Maximum a Posteriori Estimation
- Cross Entropy Loss reasoning and relation to KL divergence
- If all weights are initialized to same values then what would happen
- Dropout and what happen at training time
- Multi GPU training, parallelism, PEFT, QLoRA (mentioned in my resume)
Round 3 (30/04/2025)
- In-depth Q&A of my preferred project (Tip: if any project with model building from scratch then discuss that)
- How does ViT work?
- We want to build a visual search system for Meesho? How will you approach it?
- Focus on model building
- Which model will you select and why?
- How do you train it
- Evaluation and Business Metrics
Later that day I received confirmation from HR.
Questions asked (30)
- 01
What is p-value?
Other
What do you understand by p-value?
- 02
Null vs Alternate Hypothesis
Other
If we fail to reject the null hypothesis then do we accept an alternate hypothesis?
- 03
Sample Uniformly within Unit Circle
Data Structures & Algorithms
We want to sample values between [0, 1] uniformly but it should remain within the unit circle. How do you do?
- 04
Explain Bias-Variance Tradeoff
Other
Explain bias-variance tradeoff
- 05
Address Stagnant Validation Loss
Other
Training on a large dataset, training loss is getting reduced but validation loss is stagnant, how will you address it?
- 06
Max Population Throughout History
Data Structures & Algorithms
Given a list of citizens in a country, Info of birth date and death date, Find the max population throughout the history
- 07
Max Installable Libraries with Prerequisites
Data Structures & Algorithms
Find the max number of libraries that can be installed from ‘libaries_needed’:
You are given three lists:
Libraries_needed = [‘pandas’, ‘numpy’]
Libraries_wd_no_preq = [‘A’, ‘B’, ‘C’]
Libraries_wd_preq = [[‘A’, ‘B’], [‘B’]]
Here 0th index elements are prerequisites for ‘pandas’ and so on - 08
Explain ReLU Activation Function
Other
Explain ReLU activation function
- 09
Drawback of ReLU
Other
What is the drawback of using ReLU
- 10
Explain Positional Encoding in Transformers
Other
Explain positional encoding in transformers
- 11
Sine/Cosine vs Binary for Positional Encoding
Other
Why did authors use sine and cosine? Why can’t we use binary values?
- 12
Explain Layer and Batch Normalization
Other
Explain layer and batch normalization
- 13
Convex and Non-Convex Loss Functions
Other
Convex and non-convex loss functions
- 14
Discuss CNNs
Other
CNNs (projects are CV based)
- 15
Discuss Pooling Techniques
Other
Discussion on pooling techniques
- 16
Segmentation Basics and Pooling Techniques
Other
Segmentation basics and pooling techniques in that
- 17
Classification Metrics (ROC-AUC vs PR-AUC)
Other
Classification metrics: class imbalance, comparison between ROC-AUC and PR-AUC
- 18
Possible Reasons for NaN Loss
Other
If loss is NaN then possible reasons?
- 19
Bias-Variance Tradeoff
Other
Bias-variance tradeoff
- 20
Regularization Techniques
Other
Regularization techniques
- 21
Why L1 Regularization Creates Sparsity
Other
Why does L1 regularization create sparsity?
- 22
Optimal Batch Size and Reasoning
Other
Optimal batch size and why?
- 23
Why Move in Negative Gradient Direction
Other
Why do we go into the negative of the gradient?
- 24
MLE vs MAP Estimation
Other
Maximum Likelihood Estimation and Maximum a Posteriori Estimation
- 25
Cross Entropy Loss and KL Divergence
Other
Cross Entropy Loss reasoning and relation to KL divergence
- 26
Consequences of Same Weight Initialization
Other
If all weights are initialized to same values then what would happen
- 27
Dropout During Training
Other
Dropout and what happen at training time
- 28
Multi-GPU Training, Parallelism, PEFT, QLoRA
Other
Multi GPU training, parallelism, PEFT, QLoRA (mentioned in my resume)
- 29
How does Vision Transformer (ViT) work?
Other
How does ViT work?
- 30
Design a Visual Search System for Meesho
System Design
We want to build a visual search system for Meesho? How will you approach it?
- Focus on model building
- Which model will you select and why?
- How do you train it
- Evaluation and Business Metrics