Traditional restaurant search systems, which focus mainly on cuisine genres or individual dish names, may not satisfy some users who want to enjoy a variety of foods in a single meal, or who are looking for a more nutritionally balanced diet in the same restaurant. To address this issue, this study aims to develop a restaurant search system that receives desired ingredients from the user and finds restaurants that serve menu items containing all of these ingredients at the same location. The proposed system consists of three parts: (1) a vector database that stores records of ingredients with their nutritional components, (2) a graph database that stores records of restaurants with the menu items served there and the ingredients contained in the menu items, and (3) a search system that enumerates restaurants capable of providing menu items that contain all the specified or alternative ingredients. The search system first asks the vector database to suggest ingredients alternative to each of the specified ingredients; then expands the list of specified ingredients into multiple alternative lists where each specified ingredient is replaced by alternative ingredients; and finally asks the graph database to enumerate restaurants where all of the ingredients in any of the specified or alternative lists are available at the same location. In this way, users can find restaurants that serve menu items containing all of their desired ingredients or alternatives.