Contact Center Fatigue Scoring Using Baseline Work Pace Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Agent fatigue in contact centers is poorly understood, leading to reduced performance and customer satisfaction, particularly in round-the-clock operations, as supervisors lack timely insights into fatigue levels and appropriate corrective actions.
Innovation Solution
An automated system using machine learning algorithms, specifically neural networks, models non-fatigued agent work pace and generates real-time fatigue scores, comparing them to thresholds to trigger corrective actions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If automated fatigue monitoring is implemented, then agent fatigue can be identified timely, but system complexity increases
Solution Approach 1:
The patent introduces a neural network model as an intermediary between raw work pace data and fatigue assessment. The model transforms multiple work pace parameters into a comprehensive fatigue score, enabling accurate fatigue identification while managing system complexity through automated processing rather than manual monitoring of each parameter.
Solution Approach 2:
The patent replaces manual fatigue assessment with an automated machine learning-based monitoring system. The neural network automatically processes work pace data and generates fatigue scores, eliminating the need for manual supervision and reducing operational complexity despite increasing initial system setup complexity.
2Loss of time
If real-time fatigue scoring is performed, then corrective actions can be triggered timely, but computational resources increase
Solution Approach 1:
The patent performs preliminary actions by training the neural network model in advance using historical work pace data. This pre-training allows the model to make rapid fatigue assessments in real-time without requiring complex computational resources during actual monitoring, as the heavy computational work has already been completed during the offline training phase.
3Measurement precision
If multiple work pace parameters are monitored, then fatigue assessment accuracy improves, but data processing complexity increases
Solution Approach 1:
The patent merges multiple work pace parameters into a single comprehensive fatigue score through the neural network model. By combining various parameters such as talk time, hold time, and afterwork metrics into one unified assessment, the system achieves accurate fatigue measurement while simplifying the processing complexity through integrated analysis rather than separate handling of each parameter.
Data Source
AI summary
A method for tracking a level of fatigue among contact center agents for generating corrective actions. The method includes: generating, via an automated modeling process, a baseline model for an agent that models a non-fatigued pace of work for the agent; generating, via an automated fatigue scoring process, fatigue scores for the agent in relation to successively occurring respective subperiods occurring within a current shift; and monitoring, via a fatigue monitoring process, the fatigue scores of the agent generated during the current shift for triggering an automated action. The automated modeling process may include identifying intervals within respective shifts of the agent for representing non-fatigued pace of work data for the agent; and creating a training dataset from measurements taken during each of the identified intervals via a monitoring script values associated with a plurality of work pace parameters.