There are many type of model development for decision support system (DSS), but the uncertainty of their outcomes needs to be estimated when they are utilized for decision support. We review various methods that have been or could be applied to evaluate the uncertainty related to deterministic models outputs and proposed our new model which is a hybrid model in order to improve from two different model that proposed by previous researcher. The hybrid of knowledge based model and old customarily used model are coming with seven phases of analysis. The phases are problem identification, problem analysis and synthesis, available alternative generation and solver, model development, alternative analysis, choices and executions. Building a simple, easy-to-apply model but with a complete phases can help in explaining and describing the process of development for the decision support system. The hybrid knowledge-based decision support system architecture model to suite with the contemporary and future business management.