PurposeThis study aims to develop a software agent-knowledge procurement management tool to address uncertainties from external and internal environments, such as record failure, slow logistics auditing and distribution delay toward improving procurement management in retailing enterprises.Design/methodology/approachQuantitative methodology was used to collect data using a self-administered survey from randomly selected procurement staffs, marketers and customers to measure their perception regarding the feasibility and acceptance of the implemented agent-knowledge-based procurement management tool.FindingsResults from empirical analysis reveal that the implemented tool facilitates collaboration and interaction among buyers, sellers and procurement managers toward enhancing procurement managers’ flexibility to handle unexpected exceptions. In addition, results confirm the feasibility of the implemented tool in supporting procurement management toward handling inventory failure exception, which occurs in traditional procurement approaches. Moreover, descriptive results from user acceptance test verify that the tool was accepted by the respondents.Research limitations/implicationsThe limitation of this study is that the implemented tool is evaluated using data collected from respondents in Malaysia retailing enterprise only; thus, the results cannot be generalized to other enterprises and country. In addition, research implications from this study design a methodological and comprehensive software agent-knowledge-based model that support buyers, sellers and procurement managers with information to facilitate buying and selling operations.Practical implicationsPractically, the designed software agent-knowledge-based model describes how software agents collaborate with each other to facilitate procurement tasks and also use the knowledge base in the implemented tool to provide information sharing platform that manages the dynamics of procurement operations.Social implicationsThis research integrates software agents which are autonomous programs that carryout pre-defined task on behalf of end users. Socially, this study would be useful for procurement managers in developing mechanisms for instilling insights in retailing operations.Originality/valueThis research is among the first to attempt to develop a software agent-knowledge-based model to support procurement management in the retailing enterprise domain. It contributes to promote e-procurement practices by implementing a software agent-knowledge-oriented tool to address uncertainties experienced in retailing enterprise. It is envisaged that this study will provide basis for future research into e-procurement practices for retailing businesses in Malaysia and beyond.