The Vitruvius platform is focused on vehicles and the possibility of working with their multiple sensors, and the real-time data they can provide. With Vitruvius, users can create software applications specialized for the automotive context (e.g., monitor certain vehicles, warn when a vehicle sensor exceeds a certain value, etc.), with the help of fuzzy rules to make decisions. To create applications, users are provided with a domain-specific language that greatly facilitates the process. However, drivers and some passengers cannot create applications on the fly since they need to type to accomplish such a goal. In this paper, we present an adaptive speech interface to allow users to create applications by only using their voice. In addition, the application is based on fuzzy rules to suit the level of experience of users. The application provides an interface that is balanced between the amount of work users have to do and the help the system provides based on the knowledge and ability of each potential user.