PayTm or ‘Pay through mobile’ is India’s multinational company that specialises in digital payment, financial services and e-commerce.
PayTm extends the value proposition for customers by gleaning insights from customer data to recommend best services, products and cross-sell financial and nonfinancial products.
Therefore, Data Scientists are in hot demand. The interview rounds for a Data Scientist role are a little different than the rest of the companies because the interview begins with an on-site visit at the PayTm office.
How many interview rounds are there in PayTm?
Once your application is selected, the recruiter will contact you and fix a date and time for the on-site interview. After which you will be subjected to 4 loops of interviews which will be 2 technical rounds, 1 round with the manager and 1 round with the HR.
PayTm Data Science Interview Process
1st Loop: Technical Screening
This round is mainly concerned with judging your technical abilities as a potential candidate. You will be subjected to SQL and Coding related questions like:
- Can you rate yourself in SQL out of 10? (Followed by basic SQL questions) What is a unique key?
- You have 2 tables. A and B. Table A has 1 column X having some values and nulls. Table B has 1 column Y having populated and null values. Join these two tables based on column x and y and count the number of output records for all the joint conditions.
- You are given an array and you need to find pairs whose sum equals to the given value x.
2nd Loop: Technical Screening
The second technical round is a little different as it correlates with your past work experiences. Questions about previous projects, your approach, the kind of data used, kind of problem faced, are asked. Following this, a favourite but very common bank of questions are asked from Big data questions in relation to Apache Spark. They somewhat go like:
- What is spark?
- What is rdd?
- How does driver work?
- How does spark work in cluster mode?
- Why are rdd’s fault tolerant?
- What is lazy evaluation?
- What is a lineage graph?
- What is persistence caching in spark?
- How can you optimize your spark application?
A coding question that might also be thrown your way is:
- You are given a singly linked list and you need to reverse it.
Pro Tip: If you are well versed in DSA, you can solve this easily. So it’s good to brush up on that.
3rd Loop: Interview with the Hiring Manager
This round is purely based on managerial related questions. No in-depth technical questions are asked. This round concerns itself with behavioural and situational based questions that judge your ability to solve problems and deal with a multitude of situations within a time crunch.
Some of them go like:
- How would you resolve any conflict with your co-worker?
- How well do you follow strict timelines?
- How should you deal with a disagreement between you and your manager?
4th Loop: Round with the HR
After successfully flying past all the previous rounds, one gets to sit with the HR to talk about next steps. Basic questions on the lines of background, resume are asked and a proper discussion of the job role takes place. You will also be acquainted with the department you will be an addition to.
20 PayTm Data Scientist Interview Questions
Here are 20 questions you must practice before appearing for the PayTm Data Scientist interview.
- How is logistic regression done?
- How do you find RMSE and MSE in a linear regression model?
- What is the significance of p-value?
- For the given points, how will you calculate the Euclidean distance in Python?
- In your choice of language, write a program that prints the numbers ranging from one to 50.
- Differentiate between univariate, bivariate, and multivariate analysis.
- Explain the steps in making a decision tree.
- How do you build a random forest model?
- How can you select k for k-means?
- How can outlier values be treated?
- How can you calculate accuracy using a confusion matrix?
- What are recommender systems?
- You are given a data set consisting of variables with more than 30 percent missing values. How will you deal with them?
- What are the feature selection methods used to select the right variables?
- Write the equation and calculate the precision and recall rate.
- ‘People who bought this also bought…’ recommendations seen on Amazon are a result of which algorithm?
- How can time-series data be declared as stationery?
- What are dimensionality reduction and its benefits?
- How will you calculate eigenvalues and eigenvectors of the following 3×3 matrix?
- How should you maintain a deployed model?
Since this Indian FinTech company has come into existence, applications of digital payments have skyrocketed and similarly the demand for Data Scientists at PayTm have increased as well. Hopefully you have gained enough insight to successfully crack a Data Scientist interview at PayTm.