Abstract. Industrial Training, especially training geared towards Industry 5.0 – referring to robot and smart machines working alongside people, is an evolving field, and recent technological advancements in Extended Reality (XR) and Arti- ficial Intelligence (AI) have propelled interest toward this goal. The combination of these technologies allows the implementation of immersive, adaptive, and per- sonalized learning experiences, which can be utilized by the workforce in on- and off-the-job contexts to address training in increasingly complex industrial systems. However, the adoption of XR-based training faces several challenges, including computational demands, latency, usability constraints, and personalization. To address these limitations, the XR5.0 Training Platform provides a state-of-the-art cloud infrastructure and AI-enhanced training solution designed to create, man- age, and display XR content to users with optimized performance and accessibility. The platform is structured around three (3) core components, namely: (i) the Holo- light Hub, for managing and orchestrating XR applications enabling low-latency streaming via a cloud-based infrastructure; (ii) the XR Training Asset Repository to ensure secure storage of training materials; and (iii) the XR Training Man- agement System for the creation, management, and visualization of XR-native training programs. This platform addresses the limitations of existing training platforms while reducing hardware dependency by adopting a device-agnostic approach. This ensures a more efficient and scalable training ecosystem, enhanc- ing workforce alignment with Industry 5.0 environments. This paper presents the platform’s architecture, key functionalities, and integration strategies while dis- cussing its potential to transform ind