Recently, the use of traditional medicines for medical treatment and maintaining good health is increasing. This condition has been followed by the increasing number of research activities, publications and databases of crude drug systems to support the scientific aspects of traditional medicines. Nevertheless, the information about traditional medicines is scattered in an unorganized manner and the existing traditional medicine databases have been stored by different database schemas. Therefore, a standardized tool that integrates and provides information about various herbal medicines is needed. In this study, we developed a mobile application with Waterfall Method, called as Herbal Medicine Systems, for various kinds of crude drug systems as a reference of traditional medicines. The KNApSAcK Family Databases were used as a role model to design the database schemas of crude drug systems by using data warehouse pre-processing technique. The Herbal Medicine Systems application is preloaded with 336 and 5,310 Kampo and Jamu formulas respectively from the KNApSAcK Kampo and KNApSAcK Jamu databases. In addition, the Herbal Medicine Systems can predict the efficacy of crude drug combinations by using Random Forest classifier developed based on the formulation of Indonesian Jamu with the average accuracy of 90%.