E-mental health applications have gained popularity recently as a result of applying developing technologies to leverage advantages over traditional care. First, e-mental health applications are inherently scalable, whereas traditional care involves one-to-one relations between patient and clinician. Second, electronic data formats can facilitate interoperability in a way that medical data on paper cannot. Third, the flexibility of a web application warrants that improvements in the care program exposed through the application will immediately benefit all applicable users. For the specific domain of e-mental health, the primary source of medical data are online questionnaires, and most applications in this domain focus on viewing, filling out, and managing these questionnaires. Many distinct such applications have been developed in recent years, most of them starting from scratch and running in isolation. In this paper, we design and propose a generic architecture for e-mental health applications from two case studies. These case studies are two e-mental health applications called HowNutsAreTheDutch and Leefplezier, and are used on a large scale with over 13,000 active users combined. By abstracting functionalities into reusable Service-Oriented Architecture interfaces, we can maximize data interoperability while minimizing application-specific code to facilitate rapid development of e-mental health applications.