The Army’s Training and Doctrine Command (TRADOC) has described plans for modernizing Army training in documents such as the Army Learning Model (ALM, TRADOC PAM 525-8-2). The ALM calls for increasing the personalization of the soldier learning process so that training is tailored to the individual soldier throughout his/her career. To accomplish this goal, a persistent representation of soldier performance across a variety of technologybased training systems is required. Currently, performance data throughout the live, virtual, constructive, and gaming (LVCG) spectrum is not maintained, nor is it used to adapt future training for soldiers or their units. However, advances in data interoperability have recently made development of complex student models using this performance data a possibility. The Experience API (xAPI) is one such innovation. As part of our research, we have used the xAPI to capture interoperable performance data for unstabilized gunnery simulators. Using this performance data, we have developed an adaptive training curriculum in which crew training is adapted based on prior individual performance on a gunnery simulator. This paper describes the development of interoperable performance data for unstabilized gunnery simulators using the xAPI specification as well as the findings of an experiment to demonstrate gains in learning and training efficiency. The results can be used to inform the Army in its training modernization goals, as well as the simulation-based training community as a whole.