Preparing for exams is often a stressful process for students. In order to better support this process, an innovative system was developed as part of the 2LIKE project that enables individualised exam preparation tailored to the specific needs of each student. The system organises the questions according to the sub-topics of a learning unit. Students receive a randomised question for a sub-topic that corresponds to their level of knowledge. If the question is answered correctly, the system moves on to the next topic. However, if an incorrect answer is chosen, the system provides feedback and generates a prompt that can then be run in a large language model, such as ChatGPT or Bing AI, to find further appropriate learning materials and resources. These offer students the opportunity to delve deeper into the topic. By repeatedly working on questions from the same topic, students can continuously improve their knowledge and skills and, thus, better prepare for their exams.