Product bundling refers to the combination of several products for sale as one product. Current bundling approaches lack the ability to adapt, focusing mostly on the creation of pre-computed static bundles, which prove to be inefficient considering the dynamically changing prospective customer needs and product availability, as is particularly the case in enterprise networks. This paper describes a novel approach for dynamic generation of personalized, constrained and rule-based product bundles in such environments. The proposed agent-based approach involves estimation of substitution and complementarity associations between products and constructing bundles according to individual customer preferences. The process adapts automatically to changing circumstances, such as customer profile, product availability and constraint and rule diversity. The proposed approach is discussed in the context of e-Furniture, an agentbased system supporting networking of furniture and wood product manufacturing enterprises.Index Terms-Adaptive product bundling, constraint-based bundling, rule-based product bundling, enterprise networks.