Recently, with growing public use of large language models (LLMs), researchers and educators have begun thinking about the role of LLMs in learning. Conversational Intelligent Tutoring Systems (CITSs) powered by Generative AI (GenAI) have added a new dimension to educational technology, characterised by enhanced contextual understanding, real-time content generation, and dynamic feedback elaboration. These systems are paving the way towards a broader adaptive education, surpassing the limitations of conventional educational approaches. Despite these capabilities, several important challenges remain unaddressed, such as understanding how learner personalisation can be systematically and quantifiably designed to scaffold learning supported by CITS and LLMs. Thus, the main purpose of this study is to design, implement, and evaluate a framework for personalised CITS that leverages GenAI to accommodate learner variety in personality traits in learning. The proposed CITS will be tested with learners, focusing on introductory programming and assessing the impact of personalisation on knowledge retention and learners’ experience. Furthermore, the research aims to formulate a set of performance metrics to evaluate the effectiveness of CITSs in conversation quality and learner engagement.