Adaptability and self-directedness are important competences of military personnel that are required for life-long learning in a constantly changing operational environment. A large research program for the Dutch Ministry of Defense, called "Education and individual training in a dynamic operational context” (2017 – 2021), investigates how training innovations can enhance such requirements. One of these innovations is personalized learning. This involves adapting training to the needs of military personnel, so that they can learn just in time and just enough. Consequently, learning will be maximally efficient and effective. Interventions like task-selection, feedback, instruction, and scaffolds, are adapted to learners’ personal learning profiles. Another such innovation is learning analytics, which involves the collection, analysis, and interpretation of data to improve learning. Learning analytics and big data technology allow for constructing better and more refined learner profiles. Learning analytics can provide new opportunities for determining appropriate, personalized learning paths. Two experiments are presented. In the first experiment, an experimental group that receives personalized maintenance training is compared to a control group that receives standardized training. The second experiment focuses on learning analytics, implemented in the TACTIS simulator CV90 shooter training. Various types of learner data are combined into learner profiles: simulator log data, instructor assessments, self-ratings on academic/psychological factors (e.g., motivation, effort), and demographics. Advanced analysis techniques are used (i.e., time-series analyses and growth curve analysis). With such data-driven learning profiles, learners and instructors can monitor learning processes over time. Ultimately, this information can be used to personalize the training.