Interview Questions

Suyash kumar T. - "I will talk about the carwash landscape here a bit. How futuristic do you want the solution to be - flying cars kind of futuristic or something that I can use even now ? What do you mean by a car wash - Is it a physical location where I take my car to OR is it something that I can install at home ? There are two sides to a car wash set up - the car wash company and the user - I am going to assume that we are building this solution for the user. Is there an specific goal in mind or I can cho"See full answer

Georgi S. - "Predicting user behavior generally is hard, unless you incorporate in all your product changes some kind of psychology based trigger or make behavior science-driven changes. For example creating scarcity like Booking.com who tell you that the rooms are about to be booked by someone else, or that others are looking at the property or booking actively today. Otherwise you'd have to rely on A/B testing and feature flags to release and measure the impact of product updates. A/B testing is a statis"See full answer

Himanshu G. - "First, I’d start by checking the alignment of each idea with our core business goals. If any idea doesn't directly contribute to those goals, I’d deprioritize or eliminate it upfront. Next, I’d use a scoring model like RICE (Reach, Impact, Confidence, Effort), especially because effort is a critical factor when resources are limited. This gives us a structured and quantifiable way to rank the ideas. Once we have a prioritized list based on scores, I’d take it a step further and evaluate key as"See full answer

John C. - "Professional, congenial people"See full answer

Andrei S. - "Follow up question was: what is your strategy to sell it?"See full answer

Anonymous Muskox - "So Machine learning provides the ability to machines to learn patterns from large data. So applying the same when needed. So the type of algorithm depends on the requirement/use case or the data. For example, if we have data that is labeled and we need to do classification then we will go and perform logistic regression but if we want prediction instead of classification, then we will go and build a regression model. In case the data is not labeled, then we can go ahead and build a model usin"See full answer

Matthew P. - "I am"See full answer

Anonymous Muskox - "Metrics which Youtube Consider before building a recommender system Number of likes on a video by user The watch time of a video by the user The video disklied by the user The video share by a user The video skipped or churn with 20-30 seconds. Depending on this Youtube build a recommender system. The video suggestion feature in youtube works based on the recommender system. It may use a hybrid of batch prediction and online prediction. So depending on the above metrics, the youtube p"See full answer

R - "Assumptions I am going to assume that users know how to send smell through the phone I am also going to assume this is only available between phone-to-phone and not server-to-phone Clarifying questions What are the goals? Customer retention or New market/user acquisition? I would like to pick New Market/User acquisition because this is a new technology that can be used in new use cases in new verticals. Metrics We will measure number of users who used this new technology We will measure"See full answer

Exponent - "This is a Measure Success question with a slight twist. The twist here is we need to consider a hypothetical product rather that one already built. This changes our formula slightly - specifically we may not be able to apply a UX flow to drive analysis since we're unsure of the implementation. Instead, we'll look at core behaviors that are indicative of success. Here's the modified formula: Ask clarifying questions State the goal of the feature **Apply a UX flow to drive a"See full answer

T I. - "This is due to sticky sessions. The load balancer is not correctly configured with sticky session option. It is likely the servers were storing session data on the server themselves (in-memory), and thus when user makes a request, the load balancer routes this to a different server than the one they started with, that second server may not recognise the user's session. This could prompt the user to log in again. One way to resolve this, is to use a centralised session storage, something like"See full answer

Erjan G. - "i said there is hashed, clustered, non-clustered"See full answer