Barclays Data Science interview questions

How you decide whether you need data science project or not?

Pick a choice from three machine learning models and describe it as if to a non-technical stakeholder. (neural net, random forest, and one more I can't remember). If you have an effective solution and it meets internal resistance, how would you get it's value across to that person? If we give you <XYZ> data, what would be your initial approach? Standard questions about technical experience: have you ever set up an API, have you used time series dependence, NLP, etc. etc. These are not dealbreakers for them though. A primary focus of both rounds is not just your technical expertise but how well you can put technical concepts into simple language.