Over the past few years, Mobility-as-a-Service (MaaS) has become an alternative to owning a private car, integrating various public and shared travel modes in a single platform. Despite its attractiveness in meeting everyday travel needs, it has not seen high penetration rates within the urban population. A solution to increase its acceptability is to improve its personalisation so that it can better cater to the needs of each user. In this paper, we propose the global architecture of a MaaS tool under development for the city of Toulouse and its surroundings. The proposed architecture combines the information about users' mobility patterns and their individual preferences along with real-time traffic updates to proactively recommend personalised routes. Within the application, users' probability of choosing a particular transport mode, calculated using discrete choice models, as well as various gamificationbased elements are integrated in order to understand users' travel behaviour and promote sustainable transportation.