Improving the Efficiency of Work Order Management by Infusing AI-Empowered Automation

Abstract

Numerous efforts in the field of AI are aimed at automat-ing repetitive tasks in business processes. In some cases, processes are reliant upon human decision-making. We hypothesize that providing AI-generated recommendations at key decision points in a business process assists in reducing the workload and errors. We implemented two predictive AI models, providing recommendations in two key stages of a work order management process. We tested the models performance for accuracy in a case study and received positive feedback from test users. In our current work (in progress),we evaluate the effectiveness of the AI recommendations us-ing an experiment, recording the performance metrics with and without the help of the models. We aim to test for the statistical significance of the assistance effect with respect to a series of variables.