This project addresses the challenge of forming high-performance, high-viability teams with significant heterogeneity and affinity levels. This project aims to create an open-source computational network model to optimize team formation by integrating functional individual differences and member affinity. The researchers train and test this computational model using empirical data from four distinct teaming contexts: scientific collaborations, open-source software development, space crews simulating long-duration space exploration missions, and project teams among students in educational and executive development programs. Utilizing the Multitheoretical Multilevel (MTML) framework, the researchers explore how teams within social networks can form at various levels (individual, dyad, triad, and group) and seek to maximize heterogeneity in certain characteristics while enhancing members’ affinity in others. Key goals include developing an MTML model to elucidate team assembly mechanisms, implementing a data-driven computational model to estimate team performance and viability based on member and network attributes, conducting virtual experiments to explore the impacts of various team configurations, and publishing an interactive web-based Exploratorium for users to simulate team performance and viability dynamically.