Upload Button Icon Add office photos
Premium Employer

i

This company page is being actively managed by Chubb Team. If you also belong to the team, you can get access from here

Chubb Verified Tick

Compare button icon Compare button icon Compare
3.9

based on 162 Reviews

Filter interviews by

Chubb Data Scientist Interview Questions, Process, and Tips

Updated 12 Feb 2024

Top Chubb Data Scientist Interview Questions and Answers

  • Q1. how will you get the embeddings of long sentences/paragraphs that transformer models like BERT truncate? how will you go about using BERT for such sentences? will you use ...read more
  • Q2. in what scenarios would you advice me to not use ReLU in my hidden layers?
  • Q3. how does backpropagation in neural networks work?

Chubb Data Scientist Interview Experiences

2 interviews found

I applied via Naukri.com and was interviewed in May 2022. There were 4 interview rounds.

Round 1 - Resume Shortlist 
Pro Tip by AmbitionBox:
Keep your resume crisp and to the point. A recruiter looks at your resume for an average of 6 seconds, make sure to leave the best impression.
View all tips
Round 2 - Technical 

(6 Questions)

  • Q1. (interviewer observed that i am nervous. so, he did some small talk, i presume to make me feel comfortable)
  • Q2. (after describing a project on my resume) explain recall in plain english? why did you use recall as evaluation metric? how do you calculate recall for mutli-label data? how did you decide the number of hi...
  • Ans. 

    Recall is a metric used to measure the ability of a model to find all relevant instances in a dataset.

    • Recall is the ratio of true positives to the sum of true positives and false negatives.

    • It is used as an evaluation metric to assess the model's ability to identify all positive instances correctly.

    • In multi-label data, recall can be calculated for each label separately and then averaged.

    • The number of hidden layers and n...

  • Answered by AI
  • Q3. What is universal approximation theorm?
  • Ans. 

    Universal approximation theorem states that a neural network with a single hidden layer can approximate any continuous function.

    • A neural network with a single hidden layer can approximate any continuous function

    • It is a fundamental theorem in the field of deep learning

    • The theorem applies to a wide range of activation functions

    • The number of neurons required in the hidden layer may vary depending on the complexity of the ...

  • Answered by AI
  • Q4. How does backpropagation in neural networks work?
  • Ans. 

    Backpropagation is a supervised learning algorithm used to train neural networks by adjusting weights to minimize error.

    • It involves propagating the error backwards through the network to adjust the weights of the connections between neurons.

    • The algorithm uses the chain rule of calculus to calculate the gradient of the error with respect to each weight.

    • The weights are then updated using a learning rate and the calculate...

  • Answered by AI
  • Q5. In what scenarios would you advice me to not use ReLU in my hidden layers?
  • Ans. 

    Avoid ReLU when dealing with negative values or vanishing gradients.

    • When dealing with negative values, use Leaky ReLU or ELU instead.

    • When facing vanishing gradients, use other activation functions like tanh or sigmoid.

    • In some cases, using ReLU in all layers can lead to dead neurons.

    • Consider the nature of your data and the problem you are trying to solve before choosing an activation function.

  • Answered by AI
  • Q6. Some questions on Named Entitiy Recognition(NER)
Round 3 - Technical 

(8 Questions)

  • Q1. (after describing a project on my resume) how did the end users use this project? why did you choose to use the techniques/methods that you di? if you don't have labelled data, how would you go about doing...
  • Q2. How will you get the embeddings of long sentences/paragraphs that transformer models like BERT truncate? how will you go about using BERT for such sentences? will you use sentence embeddings or word embedd...
  • Ans. 

    To get embeddings of long sentences/paragraphs truncated by BERT, we can use pooling techniques like mean/max pooling.

    • We can use pooling techniques like mean/max pooling to get embeddings of truncated sentences/paragraphs.

    • We can also use sliding window approach to get embeddings of overlapping segments of the long input.

    • For using BERT on such long inputs, we can use sentence embeddings or word embeddings depending on t...

  • Answered by AI
  • Q3. (after describing a project on my resume) brief all the pre-processing you have done for this multi-label text classification problem!
  • Q4. If you were an animal or bird, which one would you want to be and why?
  • Ans. 

    I would want to be a dolphin because of their intelligence, social nature, and ability to swim freely in the ocean.

    • Dolphins are known for their high level of intelligence and problem-solving abilities.

    • They are highly social animals and live in groups called pods.

    • Dolphins have the ability to communicate with each other using a complex system of clicks, whistles, and body movements.

    • They are also known for their agility a...

  • Answered by AI
  • Q5. What makes you angry/frustrated on your teammates or people at work?
  • Ans. 

    Lack of communication, unresponsiveness, and lack of accountability frustrate me.

    • Lack of communication: When teammates fail to communicate important information or updates, it hampers the progress of the project.

    • Unresponsiveness: When teammates do not respond to emails, messages, or requests in a timely manner, it slows down the workflow and creates bottlenecks.

    • Lack of accountability: When teammates do not take respons...

  • Answered by AI
  • Q6. Why do you want to join Chubb? [have a genuine answer ready, not a googled one]
  • Q7. If you manager/upper-management has asked you to do something in a way that seems stupid to you, how will you tackle this? or will you continue to do it the stupid way?
  • Ans. 

    I would respectfully discuss my concerns with my manager and propose alternative solutions.

    • Initiate a conversation with the manager to understand their perspective

    • Express concerns and provide logical reasoning for alternative approaches

    • Present data or examples to support the proposed solutions

    • Maintain a respectful and professional attitude throughout the discussion

  • Answered by AI
  • Q8. Why are you looking for new opportunities after only 1 and half years at your current company?
  • Ans. 

    Seeking new opportunities to further develop my skills and contribute to a more challenging and dynamic environment.

    • Looking for new challenges and growth opportunities

    • Seeking a more dynamic and innovative work environment

    • Wanting to work on a wider range of projects and gain diverse experience

    • Desire to contribute to a company with a stronger focus on data science

    • Exploring opportunities for career advancement and profess

  • Answered by AI
Round 4 - HR 

(2 Questions)

  • Q1. Why are you looking for new opportunities or change?
  • Ans. 

    I am looking for new opportunities or change because I want to further develop my skills and contribute to a dynamic and innovative team.

    • Seeking new challenges and growth opportunities

    • Interested in working with a diverse and talented team

    • Want to contribute to innovative projects and make a meaningful impact

    • Desire to expand knowledge and skills in data science

    • Looking for a company culture that aligns with my values and

  • Answered by AI
  • Q2. Why are trying for more offers, when you already have two offers with you?
  • Ans. 

    Exploring more offers to make an informed decision and maximize career growth.

    • To evaluate the best fit for my skills, interests, and long-term goals.

    • To negotiate better compensation and benefits.

    • To gain exposure to different industries, technologies, and projects.

    • To expand professional network and opportunities.

    • To ensure job security and minimize risks.

    • To have a backup option in case one offer falls through.

    • To explore ...

  • Answered by AI

Interview Preparation Tips

Interview preparation tips for other job seekers - interviews have been a lot practical, which i prefer to asking definitions that are just a google search away. if you do not answer to something, explain only that you know to the extent you do or just tell the interviewers that you do not know. intervieweres have also been more interested in my approach to the questions i did not have proper answers for. and interviewers didn't have a prepared set of questions, it was all spontaneous based on the projects on work, based on what I uttered. be prepared to answer and explain every word that come out of your mouth. questions may arise from anything. interviewers counter-question all that you explain, you must be able to explain why that can't be the case and why whatever that you have done makes more sense and fits the problem at hand.

be humble and calm, behaviour goes a long way.

Skills evaluated in this interview

Interview experience
3
Average
Difficulty level
Moderate
Process Duration
Less than 2 weeks
Result
Not Selected

I applied via Recruitment Consulltant and was interviewed before Feb 2023. There were 2 interview rounds.

Round 1 - Coding Test 

2 question
1. Find the highest stock proce profit in a day.
2. To rollup and aggregate claims at covergae level.

Round 2 - HR 

(1 Question)

  • Q1. Salry negotiations

Data Scientist Interview Questions Asked at Other Companies

Q1. Special Sum of ArrayYou have been given an array/list ‘arr’ of le ... read more
Q2. for a data with 1000 samples and 700 dimensions, how would you fi ... read more
asked in Affine
Q3. you have a pandas dataframe with three columns, filled with state ... read more
Q4. Clone a Linked List with random pointersGiven a linked list havin ... read more
Q5. Technical QuestionGiven a API reference. You had to make a post r ... read more

Data Scientist Jobs at Chubb

View all

Interview questions from similar companies

Interview experience
4
Good
Difficulty level
Moderate
Process Duration
Less than 2 weeks
Result
Not Selected

I applied via Naukri.com and was interviewed in Aug 2024. There was 1 interview round.

Round 1 - One-on-one 

(2 Questions)

  • Q1. Whats is supervised learning
  • Ans. 

    Supervised learning is a type of machine learning where the model is trained on labeled data to make predictions or decisions.

    • Uses labeled data for training

    • Predicts outcomes based on input features

    • Examples include regression and classification algorithms

  • Answered by AI
  • Q2. What is unsupervised learning
  • Ans. 

    Unsupervised learning is a type of machine learning where the model is trained on unlabeled data without any predefined output labels.

    • No predefined output labels are provided for the training data

    • The model must find patterns and relationships in the data on its own

    • Common techniques include clustering and dimensionality reduction

    • Examples: K-means clustering, Principal Component Analysis (PCA)

  • Answered by AI

Interview Preparation Tips

Interview preparation tips for other job seekers - do hard work

Skills evaluated in this interview

Interview experience
5
Excellent
Difficulty level
-
Process Duration
-
Result
-
Round 1 - Technical 

(1 Question)

  • Q1. Relevant question to the projects

Data Scientist Interview Questions & Answers

MetLife user image Ramachandra Maharshi Andukuri

posted on 17 Mar 2024

Interview experience
1
Bad
Difficulty level
-
Process Duration
-
Result
-
Round 1 - Technical 

(1 Question)

  • Q1. Walk us through you resume
Interview experience
4
Good
Difficulty level
Moderate
Process Duration
Less than 2 weeks
Result
Not Selected

I applied via Naukri.com and was interviewed in Aug 2024. There was 1 interview round.

Round 1 - One-on-one 

(2 Questions)

  • Q1. Whats is supervised learning
  • Ans. 

    Supervised learning is a type of machine learning where the model is trained on labeled data to make predictions or decisions.

    • Uses labeled data for training

    • Predicts outcomes based on input features

    • Examples include regression and classification algorithms

  • Answered by AI
  • Q2. What is unsupervised learning
  • Ans. 

    Unsupervised learning is a type of machine learning where the model is trained on unlabeled data without any predefined output labels.

    • No predefined output labels are provided for the training data

    • The model must find patterns and relationships in the data on its own

    • Common techniques include clustering and dimensionality reduction

    • Examples: K-means clustering, Principal Component Analysis (PCA)

  • Answered by AI

Interview Preparation Tips

Interview preparation tips for other job seekers - do hard work

Skills evaluated in this interview

Interview experience
4
Good
Difficulty level
-
Process Duration
-
Result
Selected Selected
Round 1 - Technical 

(2 Questions)

  • Q1. LSTM RNN and simple RNN difference
  • Ans. 

    LSTM RNN is a type of RNN that can learn long-term dependencies, while simple RNN struggles with vanishing/exploding gradients.

    • LSTM RNN has more complex architecture with memory cells, input, forget, and output gates.

    • Simple RNN has a single tanh activation function and suffers from vanishing/exploding gradients.

    • LSTM RNN is better at capturing long-term dependencies in sequences.

    • Simple RNN is simpler but struggles with

  • Answered by AI
  • Q2. What is lasso regression
  • Ans. 

    Lasso regression is a type of linear regression that uses L1 regularization to prevent overfitting by adding a penalty term to the loss function.

    • Lasso regression helps in feature selection by shrinking the coefficients of less important features to zero.

    • It is particularly useful when dealing with high-dimensional data where the number of features is much larger than the number of samples.

    • The regularization parameter in...

  • Answered by AI
Round 2 - Technical 

(1 Question)

  • Q1. How to fine tune a llm model
  • Ans. 

    Fine tuning a LLM model involves adjusting hyperparameters to improve performance.

    • Perform grid search or random search to find optimal hyperparameters

    • Use cross-validation to evaluate different hyperparameter combinations

    • Regularize the model to prevent overfitting

    • Adjust learning rate and batch size for better convergence

    • Consider using techniques like early stopping to prevent overfitting

  • Answered by AI
Round 3 - Behavioral 

(1 Question)

  • Q1. Normal discussion
Round 4 - HR 

(1 Question)

  • Q1. Salary negotiation

Skills evaluated in this interview

Interview experience
5
Excellent
Difficulty level
Easy
Process Duration
Less than 2 weeks
Result
Selected Selected

I applied via Recruitment Consulltant and was interviewed in Feb 2024. There was 1 interview round.

Round 1 - Technical 

(3 Questions)

  • Q1. Random Forest Hyperparameters
  • Q2. Why deep learning is used over statistical models
  • Ans. 

    Deep learning is used over statistical models for complex, non-linear relationships in data.

    • Deep learning can automatically learn hierarchical representations of data, capturing intricate patterns and relationships.

    • Statistical models may struggle with high-dimensional data or non-linear relationships, where deep learning excels.

    • Deep learning can handle unstructured data like images, audio, and text more effectively tha...

  • Answered by AI
  • Q3. How XGB is better than RF
  • Ans. 

    XGB is better than RF due to its ability to handle complex relationships and optimize performance.

    • XGB uses gradient boosting which allows for better handling of complex relationships compared to RF

    • XGB optimizes performance by using regularization techniques to prevent overfitting

    • XGB is faster and more efficient in training compared to RF

    • XGB allows for parallel processing which can speed up computation

    • XGB has been shown...

  • Answered by AI

Interview Preparation Tips

Interview preparation tips for other job seekers - Clear with Fundamentals

Skills evaluated in this interview

Interview experience
5
Excellent
Difficulty level
-
Process Duration
-
Result
-
Round 1 - Technical 

(1 Question)

  • Q1. Relevant question to the projects

Data Scientist Interview Questions & Answers

MetLife user image Ramachandra Maharshi Andukuri

posted on 17 Mar 2024

Interview experience
1
Bad
Difficulty level
-
Process Duration
-
Result
-
Round 1 - Technical 

(1 Question)

  • Q1. Walk us through you resume

Chubb Interview FAQs

How many rounds are there in Chubb Data Scientist interview?
Chubb interview process usually has 3 rounds. The most common rounds in the Chubb interview process are Technical, HR and Resume Shortlist.
How to prepare for Chubb Data Scientist interview?
Go through your CV in detail and study all the technologies mentioned in your CV. Prepare at least two technologies or languages in depth if you are appearing for a technical interview at Chubb. The most common topics and skills that interviewers at Chubb expect are Machine Learning, NLP, Python, Deep Learning and Problem Solving.
What are the top questions asked in Chubb Data Scientist interview?

Some of the top questions asked at the Chubb Data Scientist interview -

  1. how will you get the embeddings of long sentences/paragraphs that transformer m...read more
  2. in what scenarios would you advice me to not use ReLU in my hidden laye...read more
  3. how does backpropagation in neural networks wo...read more

Tell us how to improve this page.

People are getting interviews through

based on 2 Chubb interviews
Job Portal
50%
50% candidates got the interview through other sources.
Moderate Confidence
?
Moderate Confidence means the data is based on a sufficient number of responses received from the candidates
Chubb Data Scientist Salary
based on 45 salaries
₹8.5 L/yr - ₹23.9 L/yr
7% more than the average Data Scientist Salary in India
View more details

Chubb Data Scientist Reviews and Ratings

based on 3 reviews

3.7/5

Rating in categories

4.2

Skill development

3.7

Work-Life balance

3.6

Salary & Benefits

4.7

Job Security

3.9

Company culture

3.6

Promotions/Appraisal

3.7

Work Satisfaction

Explore 3 Reviews and Ratings
Data Scientist - I

Bangalore / Bengaluru

5-8 Yrs

Not Disclosed

Data Scientist

Bangalore / Bengaluru

4-8 Yrs

₹ 12-23 LPA

Data Scientist

Bangalore / Bengaluru

4-8 Yrs

₹ 12-23 LPA

Explore more jobs
Senior Software Engineer
230 salaries
unlock blur

₹9.2 L/yr - ₹32 L/yr

Technical Lead
198 salaries
unlock blur

₹16 L/yr - ₹46 L/yr

Software Engineer
113 salaries
unlock blur

₹6 L/yr - ₹17 L/yr

Data Scientist
42 salaries
unlock blur

₹8.5 L/yr - ₹25.3 L/yr

Senior Analyst
42 salaries
unlock blur

₹6.5 L/yr - ₹23 L/yr

Explore more salaries
Compare Chubb with

AIG

4.0
Compare

Allianz

4.4
Compare

AXA

3.9
Compare

Zurich Insurance

4.4
Compare

Calculate your in-hand salary

Confused about how your in-hand salary is calculated? Enter your annual salary (CTC) and get your in-hand salary
Did you find this page helpful?
Yes No
write
Share an Interview