Given a dataset which we cannot even see properly on hackerrank, build 2 ML models and answer insight related questions within 2 hours. It made no sense. The dataset was in raw form, had to be studied and converted into features and target, split for training/testing and report accuracy. This is not how a data science task can be completed.
Make sure to study recommendation engines, since this is the bread and butter of their work. They don't really have any interest in your work experience, like many tech interviews.
Lots of coding questions and data analysis tasks.
¿En que año de la facultad estas cursando actualmente?
What is the reason you want to work at GSK
Why do you interest in Flipkart?
Expectation that you can give clear examples of past decisions that make you a good behavioral fit with the organization. These examples are often asked in a question format about past decisions.
Price optimization
How to build a question answering system with no training data?
How would you go about calculating the porosity of this silicon wafer?
Viewing 291 - 300 interview questions