Interview Questions

Soundharya S. - "Designing a smart fridge is something we can compare with designing our smart phone, obviously which is not as handy as our mobile phones. I can design in such a way that it is operated through our mobile phone. for ex, if we are out of town instead of worrying about the condition of the items in the fridge we can stay close with just one click away, while coming to product installation then I would say instead of taking a lot of time in searching the items, showcase a display where a list of it"See full answer

Ananya M. - "Assumption: question is about people in the US, and we define frequent bar visitor as someone who visits a bar twice a week. we are defining bars as places that people go for alcoholic beverage , not for the snacks , non alcoholic beverage or sports TV population of use 330M , 80m per age group 0-20 (eliminate because drinking age is 21 and above) 60-80 (eliminate because they prefer drinks at home or sit down restaurants or are very health conscious to go drinking) 20-30 : 40m 80%"See full answer

Avi P. - "Product management deals with new product development, enhancing existing products, monitoring the performance, planning, forecasting, monetization, pricing, product launch, and marketing of a product. Personally, I would like to improve many aspects of Product management. Try to understand the Product we are developing as much as possible. Brainstorm and think about long term goals and vision for product we are responsible for. Hypothesize the Product vision and value of the product for"See full answer

Apurv M. - "Started with the clarifying questions. Then discussed the following with the panel: Goal Users Use Cases Features Priority Metrics to measure the success"See full answer

Frank A. - "Evaluate product performance metrics against projections, rationalize performance, and identify opportunities to optimize performance"See full answer

Prachi G. - "The solution produces the same result as the 'prescribed solution' yet it does not get accepted In the test results section transcript['year'] = transcript['year'].astype(str) df = pd.pivottable(data = transcript, index = 'studentid', columns = 'year', values = 'yearlygpa', aggfunc = 'mean').resetindex() df = df[(df['2021'] < df['2022']) & (df['2022'] < df['2023'])] df['average_gpa'] = df[['2021', '2022', '2023']].mean(axis=1).round(2) return df "See full answer

Divya K. - "Explain how you implemented your telemetry and observability in previous projects."See full answer

Debajyoti B. - "Museums are multi-revenue organizations. They dont have much or any profits in the process but thats becuase they dont want to either. All the money pretty much goes to operational costs required to be a well maintained place of exhibit. Lets consider the sources they currently "make" the money from- governments(national and/or state) private donors and trustees heritage organizations with their lottery fundings self channels like admission/entry fees, tickets for exhibitions, gift s"See full answer

Googlepm 1. - "Clarifying questions : 1. What is the product Hardware/software specific ? 2. Any particular analytics done or should it be covered as a part of the solution ? 3. Any resource constraints ? 4. Any timeline in particular ? In order to internationalize a product, the following steps could be done: Identify the market/regions : Based on, say, Google analytics, find the regions which has the most clicks / attention for our product Going wide vs Going deep : Is it a single region that we are"See full answer

Anonymous Sturgeon - "Backlog Grooming, Sprint Planning, Daily Scrum, Demo, Retrospective, Scrum of Scrum, Burn-down Chart, Tasks, User Stories, Velocity, Capacity, Team-size, Story Estimation..."See full answer

George P. - "North Star - Monthly Recuring Revenue No of new signup churn/retention CAC"See full answer

Shibin P. - "Effective loss functions for computer vision models vary depending on the specific task, some commonly used loss functions for different tasks: Classification Cross-Entropy Loss:Used for multi-class classification tasks. Measures the difference between the predicted probability distribution and the true distribution. Binary Cross-Entropy Loss:Used for binary classification tasks. Evaluates the performance of a model by comparing predicted probabilities to the true binary labe"See full answer

Googlepm 1. - "Since there is ample amount of data, there are various options to use them : Track travellers : Using traffic data coupled with user's GPS, in consent with the user, find the user's exact spot, this can be valuable in case of emergencies or tracking a school bus. Better ticketing : I understand that the notion is to NOT build anything with navigation, however if we have all data on the traffic, data analytics can predict if particular routes are used most, say if three connection routes a"See full answer

Anonymous Duck - "I will work with data scientist to implement anomaly detection models to flag any deviations in a business’s review patterns over time. Like if any sudden influx of positive reviews after a negative review could indicate a reputation management effort. Or Any like any bot reviews happening in quick sessions happening from different cities/geo locations"See full answer