Most of the modeling approaches used to understand organizational worker mobility are highly stylized, using idealizations such as structureless organizations, indistinguishable workers, and a lack of social bonding of the workers. In this article, aided by a decade of precise, temporally resolved data of a large civilian organization of the US Army in which employees can change jobs in a similar way to many private organizations, we introduce a new framework to describe organizations as composites of teams within which individuals perform specific tasks and where social connections develop. By tracking the personnel composition of organizational teams, we find that workers who change jobs are highly influenced by preferring to reunite with past coworkers. In this organization, 34% of all moves across temporally stable teams (and 32% of the totality of moves) lead to worker reunions, percentages that have not been reported and are well-above intuitive expectation. To assess the importance of worker reunions in determining job moves, we compare them to labor supply and demand with or without occupational specialization. The comparison shows that the most consistent information about job change is provided by reunions. We find that the greater the time workers spend together or the smaller the team they share both increase their likelihood to reunite, supporting the notion of increased familiarity and trust behind such reunions and the dominant role of social capital in the evolution of large organizations. Our study of this organization supports the idea that to correctly forecast job mobility inside large organizations, their teams structures and the social ties formed in those teams play a key role in shaping internal job change.