Estimation of Contact Center Performance Measures in Case of Overload and Chatbot Implementation

Abstract

Chatbots are natural virtual language assistants based on the artificial intelligence (AI) and machine learning. Chatbots popularity is increasing quickly nowadays. They are used for providing real-time service and implemented in the different customer support systems. Today, chatbots are used to improve customer experience in contact centers. Chatbots facilitates on-line conversations via text or text-to-speech. One of the most important problems is call center overload, and to determine the optimal number of agents required for the customer servicing in system where chatbots are used for handling similar requests from the customers. This paper analyzes the mathematical model of a contact-center with two types of customer service - chatbots and agents. Markov process that describes model function is constructed. Vital performance measures of interest are defined through the values of stationary probabilities of model states. Algorithm of characteristics estimation is suggested based on solving the system of state equations. The numerical assessment is given; the positive effect of chatbot usage for elimination of overload is shown. The constructed model and elaborated algorithms of performance measure estimation can be used for the control of contact center functioning in case of overload.