From Talents to Team

Matching supply and demand with algorithms

More Info
expand_more

Abstract

The globalizing economy with its new goods and services, knowledge spread, and competition for talent is an increasing complexity for organizations, which requires organizations to adapt more quickly. Organizations are essential to society, as people are more productive in groups. For their continuity, it is important that organizations continuously keep adapting to their environment. The new economy requires agile organizations that can quickly (co-)produce customized responses to the demands of the market. At the present moment, the hierarchical organizational structures face limits of their usefulness and are being replaced by a lean form of organization. In their place, these decentralized organizations are growing to overcome the limitations of these hierarchical structures. The coronavirus pandemic enforced a working from home policy that strongly decreased the access to work. This development stimulated organizations to move to decentralized organizing with a marketplace for team formation. The introduction of self-managing teams enables this desired dynamic and leads to an agile organization. Virtualization has further enriched and diversified these teams. These self-managing teams are useful for constructive conflict, diversity, innovation, performance gains, and synergy. The coordination of work changes with the introduction of self-managing teams. These teams in here in a strong individual choice of people, which is both a benefit and a risk for people and organizations. A digital marketplace is a supporting institution that can provide overview, insights, and access to several opportunities by combining supply and demand of teams and talents across organizational borders. Connecting workers to relevant teams is also the main performance criterion. Still, bounded rationality can result in non-optimal individual choices on the marketplace and a one-size-fits-all approach does not match the preferences of the heterogeneous workers. A filter within the marketplace is needed to decrease the obstacles of decentral coordination of activities, however, there is limited research available on this subject. This research follows the theory of design science research with the three cycles. The design objective is to support users in self-managing team formation with a recommender system and custom filtering algorithms. This is achieved with twelve design artifacts: personalization, criteria and objectives, recommender system, data engineering, ETL script, logical data model, custom filtering algorithms, data science, validation queries, hybrid filtering algorithm, DevOps, and advice on UX design and usage. The designed system was built with the requirements of the program of business demands. The constraints for the recommender system with custom algorithms are personalized recommendations for each worker and the incorporation of platform rules. The objectives are regular updates, scalable, open-source, and easy to set up.