Global and private search engines retrieve a huge amount of information within thousands of web pages. That is based on user's search criteria. Nevertheless, it is a challenge to analyze all the information retrieved to match her/his required information. Solving this challenge needs to combine information retrieval systems (IR) with text mining and artificial intelligent approaches. This paper proposes enhanced search engines to be semantic nested search which can be used in business domain. This is via integrating Ontology, Multi-agent technology, and human computer interaction (HCI) concepts. Where Ontology provides a semantic view for understanding the pages contents then Multi-agent collects synchronously the selected information from each link. The proposed semantic nested search engine (SNS) have three phases and use four agents. The SNS engine will be applied on three case studies; jobs search, ecommerce products search and the scientific conferences search. We use different accuracy measurement such as relevant results, recall, precision and F-score. Also, we compare the proposed SNS engine with Google, Yahoo and Bing general search engines. The proposed system consume time. But the run time is not bad relative to the manual searching time inside the links, i.e. overall, the proposed SNS saves users' time and effort.