Automate back office activity monitoring to drive operational excellence

Abstract

Business process outsourcing (BPO) is growing rapidly with intensive competition. BPO providers aim to deliver high quality services with low cost. One of the key drivers is to optimize human resource utilization. It is critical to monitor and measure the activities of the practitioners in order to identify inefficient workers, unnecessary waste in operations, and non-standardized operations. Today's practices to monitor and measure the human activities are manual and error-prone. Motivated by increasing the accuracy and eliminating manual efforts for monitoring and measuring human activities, in this paper we present our research work to automatically classify and time the daily activity of a practitioner. Even though human behavior variations and noises brings substantial deviations and randomness, the developed activity classifier and timer handles the variations and reduces the noise to a satisfactory extent. The pilot results demonstrate 98.18% accuracy to classify transactions into the activity taxonomy, and 91.54% accuracy to find out the transaction cycle time therefore to aggregate to the time spent on each activity. The results are highly valued by our business partners, and the tool is considered as a revolutionary solution for human activity monitoring and measurement. © Springer-Verlag Berlin Heidelberg 2012.