“Ageing? I say yes!” A Hybrid-AI Approach to develop a web application for sustaining older workers’ work-life balance

Context Promoting healthy lives and well-being has been set as a key priority within the European social agenda and is reflected in the third Sustainable Development Goal (SDG). The urgency to address this issue has been accelerated by serious demographic changes and the relentless process of ageing affecting the population across all European countries. Several studies have showed that a poor work-life balance (WLB) – defined as the inability to effectively reconcile professional and personal responsibilities - is linked to adverse health outcomes and diminished overall well-being (e.g., Lunau et al., 2014). This evidence underscores the importance to examine how specific working conditions affect the individual’s capacity to fulfil both work and non-work roles. An extensive literature has explored the determinants of WLB, classifying them into several macro-categories, as reported in several empirical reviews (e.g., Richert-Kaźmierska & Stankiewicz, 2016; Sirgy & Lee, 2018; Singh, Aggarwal & Sahni, 2023). For instance, Richert-Kaźmierska and Stankiewicz (2016) classify these determinants into three main categories: familiar conditions, working conditions, and economic aspects. Alternatively, Sirgy and Lee (2018) distinguish between personal and organizational determinants. Personal determinants include factors such as personality traits, cultural values, and attitudes towards work and family life, while organizational determinants encompass work characteristics, company policies and supporting organizational practices. Methodology: a Hybrid-AI Approach This study builds on the results of an earlier and novel survey developed and administered to older workers in Italy. The survey aimed at exploring how relational and social dynamics - especially family expectations and social capital - shape retirement plans and perceived WLB among older workers (Bramanti et al., 2025). The survey covers four thematic areas: working conditions, WLB, retirement expectations, and socio-demographic and relational characteristics. Expanding on those insights, the present work adopts a novel hybrid-AI methodological approach to analyse WLB among older workers in Italy. Rather than merely rely on survey data, it integrates econometric analysis with artificial intelligence (AI) tools - specifically, an artificial neural network model trained on statistically significant predictors of WLB. This hybrid approach allows for a more nuanced understanding of the complex interactions between individual, familial, and occupational variables influencing WLB. Expected Results As an applied outcome, we propose the development of a web-based diagnostic application. Drawing on data patterns and findings from the previous survey, the tool enables users to complete a brief assessment and receive personalized feedback regarding their WLB profile. Importantly, the application also identifies the main drivers of dissatisfaction, offering tailored recommendations for improvement and serving as a decision-support tool for policy design. Main Implications of the Research By highlighting the multifaceted and context-sensitive nature of WLB among older workers, this research advocates for the development of targeted policy instruments and personalized welfare interventions. The web-based tool offers a scalable solution to assess needs and guide resource allocation in a data-driven manner. Furthermore, the study underscores the importance of intersectoral collaborative action among stakeholders. In fact, after identifying older workers’ needs, it is crucial that that policymakers are equipped with actionable insights and the means to address them effectively.