Resource allocation is a core and important function of every workflow management system. Resource allocation consists of twomain parts namely; matchmaking and brokering that in normal cases are combined as a single functional unit. However in anintegrated mobile environment, where the resources such as mobile workforces are combinations of human workers and mobiledevices, matchmaking and brokering can no longer be coupled. In such environments, the accuracy of resource allocation isexpected to become quite low due to frequent disconnection, mobility of resources and human factors. In addition, resourceallocation must be fair towards human resources. An accurate and fair resource allocation in mobile environments, on the otherhand can not be attained without using an independent brokerage system. To guaranty the accuracy and fairness, the desired mobileworkforce brokerage system (MWBS) must act based on real-time environmental information as well as the history of the systemwhen an allocation is required. In this paper an architectural model for a fair and accurate multi-agent MWBS is proposed based onCBR-BDI agent architecture and Active Shared-Data Space coordination model.