Effective management of IT-enabled services is becoming increasingly important. These services are often delivered by networks of knowledge workers who constitute Knowledge Intensive Service Delivery Networks (KISDN). This dissertation contributes to the effective design and management of KISDN by presenting two mixed integer programming models which integrate disparate streams of research. The first model facilitates analysis and managerial benchmarking of KISDN. We focus on how the performance of such networks depends on the interaction between workflow decisions, information flow networks (IFNs) structure and knowledge management decisions. We propose that knowledge about IFNs and worker competencies can be effectively used to make workflow decisions. Our results, based on the study of different IFN archetypes, illustrate practices for effective management of KISDN. Recognizing existing IFNs, increasing randomness in IFNs, nurturing weak or performative ties depending on the archetype, assigning tasks based on effective worker competence, and selectively delaying assignment of tasks to workers can enhance business value. The second model focuses on the design of IFNs. Organizations are increasingly creating and using IFNs to transfer knowledge. However, there is limited understanding of the design of IFNs to maximize knowledge sharing. Our results demonstrate the impact of worker competency heterogeneity, number of skills supported by the firm, and time (cost) associated with knowledge sharing on the design of efficient IFNs.