In this short talk we will introduce our app “Augmented Reality for Data Science”. Augmented Reality (AR) is a versatile technology that makes abstract data more tangible. With our app, we show how, for example, a frequency distribution of flowers as well as their shape and structure can be visualized in AR as a hologram, and classified into three species using machine learning. In addition to such simple classifications, data scientists can also use the app to ask more advanced questions, such as: What happens if a flower feature has a different value, such as the color would be different? We show how to create so-called “counterfactuals” by adjusting the values using simple steps.