Abstract. Rapidly changing business requirements necessitate the adhoc composition of expert teams to handle complex business cases. Expertcentric properties such as skills, however, are insufficient to assemble an effective team. The given interaction structure determines to a large degree how well the experts can be expected to collaborate. This paper addresses the team composition problem which consists of expert interaction network extraction, skill profile creation, and ultimately team formation. We provide a heuristic for finding near-optimal teams that yield the best trade-off between skill coverage and team connectivity. Finally, we apply a real-world data set to demonstrate the applicability and benefits of our approach.