Personalized active intervention technology is mainly based on AI to realize active diagnosis and treatment and health interaction and management, such as patient oriented medication guidance, diet recommendation, exercise intervention, etc. Traditional intervention programs have shortcomings such as high dependence on manual work and lack of personalized service. iFLYHealth proposes a personalized active intervention scheme dual-driven by knowledge and data. Through continuous analysis and processing of user data, as well as continuous upgrading of AI technology and building a more dimensional intervention knowledge graph, we can achieve population-based automatic AI intervention, and finally achieve personalized active intervention for each individual with his own experience.