Digitalization of processes in the public sector is a challenging endeavor for citizens: too long or ambiguous forms and unfamiliar terms make it challenging to participate in these processes. What is needed are methods to guide citizens with IT systems through processes in the public sector in a user-centric way. Within this paper, we investigate how to develop web systems that provide citizen-centered guidance in several dimensions: explanatory content, guided navigation, contextual process support, and form completion. Our solution integrates generative AI and model-driven engineering methods to develop web applications for the public sector. This has not only advantages for citizens, but we can also enhance the system maintenance of such systems with these methods. We present the main processes of how to realize such integrated approaches, provide some examples from practice, and discuss open challenges for the approaches.