With the continuous advancement of the DRG/DIP medical insurance payment methods reform, new problems such as upcoded disease groups and cost transfer have emerged in medical insurance fund supervision. This paper deeply studied the innovative practice of Shanghai in DRG/DIP medical insurance intelligent supervision, and systematically demonstrated Shanghai's initial exploration of regulatory indicators and rule application to the construction of an intelligentwarning model based on evidence-based trajectory of disease groups. The warning model was embedded in the real supervision scenarios, the concept of clinical pathways was introduced, and big data plus AI technology was applied to detect suspicious points, the supervision personnel's dependence on medical professionals was reduced. The application of the intelligent warning model based on the evidence-based trajectory of DRG/DIP disease groups has effectively improved the scientificity, accuracy and efficiency of supervision. In the future, it is recommended to expand the scope of DRG/DIP intelligent supervision and warning, gradually optimize the evidence-based trajectory of disease groups, and continuously strengthen the hospitals' awareness of active supervision.