An optimization approach for hybrid workflows in platform-enabled private service marketplaces

Abstract

The future of work is likely to include private platform-enabled marketplaces which remain a new, fertile ground for research. This paper focuses on hybrid workflows which utilize the expertise of in-house (IH) and on-demand (OD) workers to perform complex knowledge intensive tasks under service constraints. Our primary contributions are in identifying and modeling an innovative application of OR in the new economy, proving its complexity and proposing a high-quality heuristic. We formulate a scenario-based mathematical programming model of a private talent marketplace for Knowledge Intensive Business Services (KIBS). This model considers uncertainty in task, worker, and OD marketplace characteristics. Problem complexity is illustrated. We develop a simplified queuing theory-based model of PKIBS to characterize the structure of the OD marketplace. Propositions from the queueing model inform IH workforce size, IH and OD workforce composition, and OD pricing decisions. A Tabu-based One Period Look Ahead heuristic based on propositions from the simplified queuing theory-based model is proposed. Heuristic performance is evaluated using a problem upper bound and MIP solution for small problems, and experiments based on a cyber security use case. Results illustrate the mix of OD and IH workforce and sophisticated hybrid workflows for different market conditions.