The technical progress is evolving and each year presents new growth opportunities to different people and companies. Modern production and trade management are processing terabytes of gathered statistical data to gain helpful information for different management tasks. Most common tasks that are risen for solving are forecasting the demand for a new product, knowing only that product descripting data and creating a production plan for a product on different product life cycle phases. As the amount of data growths it becomes impossible to analyse it without using modern intelligent technologies and different systems are created and introduced to support a human in his decisions. The presented chapter is oriented on bringing intelligent agent technologies for supporting a manager’s decision in such tasks as forecasting demand for a new product and defining the current PLC phase in which the monitored product currently is. The chapter presents two multi-agent systems, designed directly for solving defined tasks. First multi-agent system analyses demand short time series and product description in order to discover a connection between those two parameters. A clustering algorithm, such as k-means, is used to discover different demand patterns and a classification decision tree is used to visualize connections between demand patterns and a product descriptive data. Second multi-agent system was designed to support a manager in his decision about the PLC phase, in which the product currently is. Such information is valuable as in the maturity phase the demand is more stable and a cyclic production planning can be applied. Two different clustering algorithms were tested – Self-Organising Maps and a Gravity based Hierarchical clustering algorithm. The chapter presents a detailed description for both systems and for each agent in systems, as also brings a description of methods and techniques used to support functionality in the systems. Both systems were trained and tested using the real demand data. The results, obtained, show that presented systems can mine and bring to manager additional information in order to support a human’s decision.