Newsenselab is developing a migraine app, which quantifies migraine and predicts attacks. M-Sense has two main innovations: The first innovation is to translate physiological data into individual dynamical biomarkers that are predictive of later attacks. M-Sense learns these early warning signs and can then detect days with an increased risk of attacks. Second, we will trigger alearning process within the user. Through an intuitive interface, the migraineur gains a deeper understanding about the connection between her migraine attacks and her own biological rhythms and lifestyle habits. Identifying and changing these influencing lifestyle factors can improve her condition. This two-fold approach of machine and user learning methods gives migraine patients a data-driven support to reduce attack frequency by making attacks more predictable and being able to schedule preventive interventions.