It was okay. The interviewer was good and helpful. Also, I'm posting this from a LinkedIn post so that I can review other companies' salaries and interview questions. So it's up to you whether you believe this or not.

The ML Engineer interview at Groww took 75 minutes, consisting of a 50-minute ML case study and 25 minutes of coding.

Round 1 included a brief discussion of my ML work experience and scaling challenges, followed by two coding problems that required Python, and an ML-based case study.

The coding problems were:

• Given a list of strings, output the common prefix characters for all of them. For example, given [flow, flower], the output is flow.
• Given a function 'f' which returns a random integer between 1 and 7, build a function using this function that can output a random integer between 1 and 10 based on probability and normalisation.

Question 3 was an open-ended ML case study discussion. Suppose you are the owner of an early-stage Zomato and you have to handle legitimate and illegitimate order refunds. How would you tackle this as a business owner and as an ML engineer?

The goal was to identify all possible use cases and think critically about how to categorise the order refunds. We discussed identifying the variables, the raw training dataset, and which algorithms to use to solve this problem at a scale of over 10 million orders on a day-to-day basis. There was also a discussion on classification, K-means clustering, and label or hot encoding challenges.