Real-time Business Intelligence has emerged as a new technology solution to provide timely data-driven analysis of enterprise wide data and information. Such type of data analysis is needed for both tactical as well as strategic decision making tasks within an enterprise. Unfortunately, there is no clarity about the critical technology components that distinguish a real-time business intelligence system from traditional data warehousing and business intelligence solutions. In this paper, we take an evolutionary approach to obtain a better understanding of the role of real-time business intelligence in the context of enterprise-wide information infrastructures. We then propose a reference architecture for building a real-time business intelligence system. By using this reference architecture we identify the key research and development challenges in the areas of data-stream analysis, complex event processing, and real-time data integration that must be overcome for making real-time business intelligence a reality.