Unfortunately it seems rather difficult to talk about the job duties without revealing the actual identity of the company that I am joining in Toronto. Instead, I will try to explain what data science means in general at different companies. There are typically four types of data science jobs. Data analytics 2.0: It can be very similar to what a management consultant does, except that you use Hive, R, or Python instead of Excel, Access, or SPSS. You try to answer business- or product-related questions such as How to segment our users (by country, age, acquisition channel, usage, etc) and which segments are the best for our business? There is a sudden drop of usage in a certain country. What happened? Product data science: Mainly A/B testing. You work with a product manager to decide which features are worth testing and run the experiments. You may need to calculate yourself things like sample size, statistical significance, etc., if there is not an already established framework. Machine learning: Logistic regression, random forest, neural networks, etc. For example your company does personal lending and they need a smart credit scoring algorithm. Hardcore research: Think about self-driving cars or go-playing machines.