Self-regulated learning (SRL) is an important skill for academic performance. Still, most university students show deficiencies in one or all areas of SRL. This skill gap can negatively impact the learning process, leading to subpar outcomes or even students dropping out of tertiary education. The good news is that self-regulated learning can be taught. However, traditional learning formats such as coaching or online courses come with the limitations of either not scaling well or lacking effectiveness. Mobile digital assistants represent one promising approach to address these problems. This work presents the preliminary results of an ongoing design science research project to design the mobile digital assistant Amsl. We describe initial design knowledge in the form of meta requirements, design principles, and design decisions. In addition, we describe the first instantiation of Amsl, intending to help students become successful self-regulated learners.