The shift from an information society to a knowledge society require rapid information harvesting, reliable search and instantaneous on demand delivery. Information extraction agents are used to explore and collect data available from Web, in order to effectively exploit such data for business purposes, such as automatic news filtering, advertisement or product searching and price comparing. In this paper, we develop a real-time automatic harvesting agent for adverts posted on Servihoo web portal and an SMS-based notification system. It uses the URL of the web portal and the object model, i.e., the fields of interests and a set of rules written using the HTML parsing functions to extract latest adverts information. The extraction engine executes the extraction rules and stores the information in a database to be processed for automatic notification. This intelligent system helps to tremendously save time. It also enables users or potential product buyers to react more quickly to changes and newly posted sales adverts, paving the way to real-time best buy deals.