The global burden of mental ill health has been rising at an unprecedented rate, while the availability of digital and artificial intelligence (AI) technologies has opened a revolutionary opportunity for mental health service delivery. Existing research, however, remains fragmented along two largely separate tracks: workplace mental health at the organisational level and individual psychological healing at the personal level, AI applications are assessed primarily for technical feasibility rather than privacy protection, algorithmic bias and ethical governance, and the business models, outcome evaluation and professional referral mechanisms of digital mental health products have not been adequately aligned with scientific evidence. It is found in this study that, by systematically integrating stress science, positive psychology, workplace wellbeing, AI technology and evidence based psychological healing, a complete knowledge system and practical model of personalised mental health management can be constructed. Stress responses are governed by the autonomic nervous system and the hypothalamic–pituitary–adrenal (HPA) axis, heart rate variability (HRV) serves as a key physiological marker of stress and the flexibility of brain–body integration, cognitive reappraisal is a more adaptive emotion regulation strategy than expressive suppression, and resilience can be enhanced through blended cognitive behavioural and mindfulness interventions. The Job Demands–Resources model and perceived organisational support theory provide the theoretical foundation for institutional workplace mental health promotion, whereas natural language processing, AI emotion recognition, generative AI and personalised recommendation constitute the technical engine of digital services, subject to stringent privacy, bias and governance safeguards. An integrated AI personalised mindfulness practice workflow and four complementary business models (B2C subscription, B2B enterprise provision, B2B2C healing service platform and clinical integration) are proposed. It is argued throughout that AI functions as an auxiliary instrument rather than a substitute for professional care, and that human–AI collaboration, professional referral boundaries and evidence based validation constitute the prerequisites for sustainable development.