The discipline of People Analytics (PA), which began as an innovative Human Resources (HR) methodology focused on managing workforce dynamics, has proven to be highly effective in optimising organisational processes and fostering employee skills development. Given the background of previous research findings in cybersecurity, this paper explores the potential of PA, in the context of Air Traffic Managers (ATM), to improve the efficiency of the training process in cybersecurity. The most problematic aspect of training in cybersecurity, but also from a general point of view, in big organisations is the optimisation of the Return on Training Investments (ROTI). This indicator, at a general level, involves three essential sources of costs: (i) class composition, (ii) long-term impact of training, and (iii) direct cybersecurity training costs. According to the Kirkpatrick/Phillips ROI model, element (i) is usually the prevalent training variable, and PA is more impactful here. The advantages of PA as an application of Learning Analytics (LA) are proven. We apply it as a method to increase the effectiveness of training as a cyber-risk reduction method. By employing Machine Learning (ML) and data-driven strategies, PA can precisely identify the employee cluster most in need of training, ensuring that these interventions achieve maximum impact regarding cybersecurity improvements. The paper describes our solution’s improvements over the classic training approach and its role and relationship in optimising class composition. Thanks to the EU-funded project SEC-AIRSPACE, this preliminary presentation reports unique design characteristics adopted by the authors for the ATM world.