This paper presents the CryptoQuiz, a web-based, AI-enabled learning platform designed to support students studying Cryptography. The system allows users to access an unlimited number of automatically generated practice questions and receive personalised feedback. The application integrates a Large Language Model (LLM) to generate tailored quizzes, informative feedback, and adapt future quizzes based on each student’s performance history. To ensure alignment with the course curriculum, the system employs a customised version of Retrieval-Augmented Generation (RAG), grounding the AI-generated content in relevant educational material. The evaluation part of the paper includes the assessment of the quality of question generation, feedback mechanisms, and adaptive learning features. The study serves as a case study in applying generative AI to personalised education, offering insights into the integration of LLMs within domain-specific learning platforms.