From chat to process: A conversational agent framework for social business process management

Abstract

Traditionally, Business Process Management (BPM) approaches have been designed for repetitive, structured processes. With Volatile, Uncertain, Complex, and Ambiguous (VUCA) environments came a high context-dependence and an increased need for knowledge in processes. That led to more and more semi-structured and unstructured processes, which resulted in a shift from BPM to dynamic BPM, ambidextrous BPM, and even social BPM. Social BPM could be an answer to the rigidity of traditional BPM approaches when organizations operate in VUCA environments if context-sensitive tool support is given. In this paper, a proposition of a framework that combines AI conversational agents and social BPM is presented to help bring added agility to BPM. These agents are thought and designed to intervene by guiding, recommending best practices, and giving feedback; therefore, acting as participants in processes. Their roles go from co-constructing and structuring processes to supervising the execution of said processes through helping with task accomplishment. This framework is being developed using an exploratory approach that combines insights from the literature, a deductive design process, and evaluation through a representative use case. By combining BPM, AI, and social media, this framework could offer possibilities in these domains and help to increase agility in BPM.