Tutoring in distance education plays a crucial role in ensuring support and facilitates learning: the role could be covered by the teacher him/herself or specialized professionals, with regard to the structure of the organization. Tasks carried out by tutors ensuring individualized paths and effective learning are often very time consuming; at least they have to: - monitor, track and assess what the student does within the learning environment (i.e. pages/documents consulted/downloaded); - analyze students’ writings; - manage knowledge sources by i.e. building representations of knowledge through conceptual maps or taxonomies; - watch the network of formal and non-formal relationships as it develops within the class, foster its strengthening and supports the inclusion of all individuals; - give formative evaluation and feedback. All these activities lead to effective planning of learning paths, and aim at student centred, individualized learning. However, considering the numbers of students involved in distance learning, the possibility to fully perform the activities needed to monitor students and therefore planning accordingly is put in danger by the burden of necessary work for each student. There is thus a the need to develop a multi-agent based, intelligent tutoring system pursuing the goal of monitoring, tracking, assessing students, giving formative feedbacks and useful data to individualize learning paths: a tool able to perform these tasks will improve self-evaluation by students and give inputs for effective planning toward an enhanced quality of education and training. The software, developed on java-based technologies, multi-platform open source, foresees modules for open source pluggable platforms exhibiting software agents as web services. It will be piloted in university, adult education and vocational education and training. The impact envisaged will be on educational/training staff, developers and researchers in educational technologies, and on education and training institutions.