This paper focuses on using the cooperative neuro-fuzzy system for the effective and customised selection of entities from large and heterogeneous resources by presenting a general architecture. An experiment is carried out with the fast-moving consumer goods to prove the utility of the architecture. It is observed that most consumers go for the frequent purchase of fast-moving consumer items. Further, various brands, costs, discounts, schemes, quantities, and reviews might make it challenging. Hence, such decisions need to be intelligent and practically feasible in terms of time and effort. The paper discusses neural networks to categorise the entities, type-1 & 2 fuzzy membership functions with rules, training sets, and graphical views of the fuzzy rules and the experiment details. Besides the generic approach and experiment, the paper also discusses the work done so far with their limitations and applications in other domains. At the end, the paper presents the limitations and possible future enhancements.