Embedding Data Science in Business Decision Making Through Analytic Agents

Cloud computing has been fueling the emergence and wide use of “analytic agents”. Simply defined, an analytic agent (AA) is an algorithmic representation of how someone makes decisions in a process. Examples include algorithmic trading in Finance industry, machine learning for autonomous cars, algorithmic diagnosis in health care. This technology can have equally disruptive impact on the way business processes are designed and executed. It is not hard to predict that, in a few years, all industries will have to use AAs in their planning and decision-making processes to stay competitive. Early adopters will gain substantial competitive advantage. Some leading tech companies are already moving in the direction although it has not been established as a scientific discipline yet. There are significant benefits of using AAs in process management. They bring a discipline to process owners to define the rules, constraints and objectives of their planning decisions comprehensively, unambiguously and accurately. Therefore, they enhance interactions and communications amongst teams that contribute to a process. When developed well, AAs can manage routine parts of a business process autonomously and therefore can help business make instant decisions in planning and execution. Since AAs are well defined computable objects, their behavior can be analyzed to identify factors that impact business performance. Extensive “what-if” analyses can be conducted through AAs to answer key questions and therefore identify best actions to improve business metrics. In this presentation, we will articulate how AAs can work as building blocks for business process design through illustrative examples.