The top one health care risk is antibiotics resistance bacteremia other than cancer. The sepsis mortality reaches up to 30%. The diagnostic time delay of sepsis may cause 7.6% of increase per hour of mortality About 30-40% of physician used inappropriate antibiotics. We developed artificial intelligent (AI) MALDI-TOF mass spectrometry to fast predict antibiotics resistance bacteria and automatic detection of physiologic and laboratory data to predict sepsis and mortality risks. We simultaneously use clinical decision support system (CDSS) and referring antibiogram (patient antibiotics history dash-board) to facilitate choice of effective antibiotics. We integrated above 4 platforms to create iAMS in China Medical University Hospital (CMUH)