The practice of business mainly involves supplying quality products and services to customers who need them. A supply chain is a network of organizations and their related activities that produces value for the consumer. Emerging technologies are attempting to enable suppliers to customize the entire marketing process. Instead of delivering custom-made products at competitive prices, seller-centric markets are being transformed into buyer-centric ones.

This paper discusses the research being conducted at IBM on developing a decision support system (DSS) for various aspects of supply chain management. The support system can serve the needs of buyers, as well as suppliers, in making decisions on what to buy, what to sell, and what to promote. The architecture of the DSS is agent-based. Query-generating agents generate the queries. Based on these queries, search agents search for a set of sites and for items that closely match the queries. Matching agents then analyze the findings of the search agents in a collaborative manner, and arrive at matches between the supplies and demands. After the matching phase, filtering agents produce a short-list of better solutions for the decision maker, who then decides on the best solution.

The electronic marketplace that incorporates the DSS under discussion uses the IBM Net.Commerce toolkit. Data exchanges between supply-chain partners over the Web employ Extensible Markup Language (XML), as does integration between enterprise resource planning systems (ERPs) and Web sites. The authors discuss the details of a paper trading company that installed this electronic marketplace. In general, the quality of presentation and discussion in this paper is high.