I went through four rounds in Flipkart.

There were two phone screens and two in-person rounds.

The phone screens basically revolved around Flipkart and its business delivery model. I asked them every possible question I had in mind pertaining to Flipkart, such as their business model, what time in the future they are thinking of breaking even, how they are planning to battle Amazon, why they increased their minimum price from ₹100 to ₹500 and if that is not going to hurt people who relied on the free shipping model earlier, what technology stack they are using, and why search is a little slow.

The second phone screen had a business question: If a commodity has to be priced between ₹48,000 and ₹50,000, and the commodity is bought at ₹42,000, with no form of discount permissible below ₹48,000 as dictated by the supplier, then what do you do to keep clients from going to the offline market where the commodity is being sold at ₹50,000?

The first in-person interview question was asked by a technical architect from the catalog team, who wanted to know how unstructured data sources about catalog information from the supplier can be merged with the catalog. What kind of sanity checks need to be done? What types of duplicate prevention measures need to be avoided? What type of data store has to be designed for storing catalog information?

The second in-person interview round had another technical architect from the online marketing team, and his question was:

Flipkart has an online marketing tool at its disposal. Now, they wish to send three types of emails:

1. Mass mails
2. Promotional mails
3. User intent mails about a product they browsed or wanted to buy, or a corresponding product related to one they already bought.

They also wish to track the rate at which they send emails and whether a weekly email is being sent to customers.

Keeping all this in mind, how do you create a system that will send emails to different segments of customers based on their buying habits, browsing history, likes and dislikes, etc., plus the data about customers provided by the business development managers?

The first in-person interview question was asked by a technical architect from the catalog team. They wanted to know how unstructured data sources about catalog information from the supplier can be merged with the catalog.

What kind of sanity checks need to be done? What types of duplicate prevention measures need to be avoided? What type of data store needs to be designed for storing catalog information?