With the increase of rehabilitation costs in the last years, tele-rehabilitation systems are becoming popular. These systems allow patients to exercise at home, without the need to commute to a rehabilitation center. In this paper we introduce a web-based platform to support and manage the entire cycle of a rehabilitation program that can be carried out autonomously at home. In particular, the platform lets therapists schedule a set of exercises and tune their difficulty level, according to patient’s residual capabilities. Moreover, it provides clinicians with the trend over time of specific indexes (accuracy, speed, amplitude, strength), computed over the whole pool of exercises carried out daily. Therapists may review patients’ activities through an avatar animated with previously acquired motion data. The platform is currently being tested on patients suffering from multiple sclerosis; however, it can be used also in other rehabilitation domains.