Abstract: This paper presents an interactive team collaborative learning and problem-solving (ITCLP) framework for effective teamwork learning and assessment. Modeling the dynamics of a collaborative, networked system involving multimodal data presents many challenges. This framework incorporates an Artificial Intelligence (AI), a Machine Learning (ML) and computational psychometrics (CP) based methodology,system architecture, and algorithms to find patterns of learning, interactions, relationships, and effective teamwork assessment from a collaborative learning environment (CLE). Collaborative learning may take place in peer-to-peer or in large groups, to discuss concepts, or find solutions to real-time problems or working on situational judgement task (SJT). Intelligent Tutoring Systems (ITSs) have been mostly used as a supportive system for the varied needs of individual learners. The ITCLP framework enables development of ITSs for team tutoring and facilitates collaborative problem solving (CPS) by creating interactions between team members. Our team model maps team knowledge, skills, interactions, behaviors, and shared knowledge of team tasks, and performance. We will collect the team interaction log data, user eye tracking, and user portrait video/audio and will map team skills evidence based on CPS, a broad range of cross-cutting capabilities, which is part of an even broader Holistic Framework (HF) proposed by Camara and colleagues.