Practical "hands-on" activities have been advocated by statistics educators as necessary precursors to technology-enhanced learning. For example, a typical learning trajectory for simulation-based inference could involve students using physical objects such as coins, cards or spinners first before moving to the use of technology to carry out simulations. In small class settings, these interactive practical activities enable rich learning conversations between the instructor and their students. However, when transforming "small-scale" activities to the large-enrolment course, additional strategies are needed to continue learning conversations between the instructor and up to 500 students. Here we present examples of large-scale interactives developed for our large introductory statistics and data science course and describe the underlying framework used to design such interactives.