In vertical search engine research, many works have been reported. But most of them focus on its key issues such as crawling, extraction, and query and few of them give a total solution for building a complete vertical search engine from scratch in a systematic method. To address this issue, we propose a model-driven method and its supporting tool SESQ. Based on a user defined ER schema for a target domain, the tool can help to build a complete search engine by integrating tasks of crawling, extraction, data management and query within one unified framework.