Learning Analytics has emerged as an active area of research fueled by the abundance of data generated by learners, teachers and educational applications in a variety of contexts. It relies on an amalgamation of data science and HCI methods to collect and process this data, predict and infer knowledge from it, communicate this knowledge back to the stakeholders of the educational process and help them interact with it effectively. Our Learning Analytics research is carried out in several directions. We are working on (dynamic) psychometric models for measuring, understanding and optimising learning, explore and evaluate the quality of collected data, and study learning environments. Several of these initiatives are with close collaboration with both experts in the field and key organisations, such as Cito, ACT and ProWise. We are developing interfaces for learning dashboards and evaluating these dashboards with real students and teachers. We also co-organize a special interest group on learning analytics, which organises occasional meetings with researchers, students and industry.