Electronic lexicography and computer-assisted language learning (CALL) have become standard ingredients of current foreign language learning and teaching (FLT). Although the two fields have a lot in common, especially now that vocabulary is slowly but surely taking centre stage in FLT, they still very much lead separate lives. And yet, a look at the respective literature shows that they face a large number of common challenges. In my presentation I will focus on one of these, i.e. the need to individualise both content and user interface to meet learners’ needs. Using notions borrowed from both CALL (adaptive CALL, learner modelling) and electronic lexicography (multifunctional dictionaries, leximats), I will make a case for the design of customizable reference-cum-learning tools and highlight the key role played by learner corpus data in the identification of learner profiles. I will present the Louvain English for Academic Purposes Dictionary, a web-based dictionary-cum-CALL tool designed to help non-native speakers of English write academic texts. The tool is characterized by its dynamic interface which automatically adapts the entry to the learners’ discipline and mother tongue background and contains both generic and L1-specific exercises.