Abstract
<title>Abstract</title> <p>The rapid growth of online services and users in today's digital society has intensified cybersecurity challenges, highlighting the need not only for technical safeguards but also for effective education and expert training to enhance overall cyber resilience. Existing cybersecurity training methods, ranging from passive learning to hands-on exercises, gamified scenarios, and cyber ranges, offer valuable opportunities but face persistent limitations in learner assessment, feedback scalability, and personalization. This paper introduces an innovative profiling model integrated within a~cyber range environment to address these challenges by continuously evaluating learner performance during hands-on cybersecurity training. Supported by artificial intelligence, the proposed model captures contextual and performance data to construct individualized learner profiles, which are subsequently used to generate descriptive automated feedback. This includes performance summaries, identification of strengths and weaknesses, and personalized recommendations. Additionally, the system derives adaptive learning paths that align training difficulty and content with each learner's evolving needs, thereby facilitating continuous development while supporting instructors with aggregated insights. A case study demonstrates the technical feasibility of our approach, showing that automated profiling and feedback can provide structured, individualized assessments for learners and aggregated summaries for instructors. By addressing a gap in cybersecurity education, this work contributes a data-driven method for objective learner assessment and personalized training.</p>