Automated Agent Response Review System for Contact Centers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current contact center systems lack an efficient method to determine which responses from agents require supervisor review, often relying on statistical sampling that may overlook critical processes used to generate responses, leading to potential errors or missed reviews.
Innovation Solution
An automated system that assesses the process used by agents to generate responses, utilizing a process database and tracking application to select responses for review by the supervisor, ensuring that both the content and context of the response are evaluated for quality assurance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If statistical sampling is used to select responses for review, then supervisor time is reduced, but the reliability of quality assurance deteriorates as critical responses may be overlooked
Solution Approach 1:
The patent replaces the manual mechanical process of supervisor review with an automated computer-based system that uses machine learning models and algorithms to evaluate agent responses, thereby reducing supervisor time while maintaining or improving review reliability through consistent, data-driven evaluation
Solution Approach 2:
The patent introduces an intermediary automated review system that acts as a mediator between agent responses and supervisor review, pre-evaluating responses using multiple criteria and only flagging those that require supervisor attention, thus optimizing supervisor time while ensuring reliable quality control
2Reliability
If automated process assessment is implemented to evaluate agent responses, then quality assurance improves by considering both content and process, but device complexity increases
Solution Approach 1:
The patent segments the quality assurance evaluation into multiple independent components including separate machine learning models for different evaluation criteria (content quality, process adherence, compliance), allowing each component to be developed, maintained, and optimized independently while collectively providing comprehensive assessment
Solution Approach 2:
The patent creates a universal automated evaluation system that handles multiple evaluation functions through a single integrated platform, including content assessment, process verification, compliance checking, and prioritization, thereby managing complexity through consolidation rather than proliferation of separate systems
Data Source
AI summary
A method and system automatically determines when to review an agent response. The method performed by an oversight device includes receiving first data including a request type of a request received by a contact center. In response to the first data, the method includes automatically determining rules corresponding to the request type, the rules indicating required steps performed by an agent for the request type. The method includes receiving second data corresponding to a response process used by the agent in generating a response to the request, the response process including performed steps performed by the agent for the request. In response to the second data, the method includes automatically determining whether the required steps are included in the performed steps of the response process. The method includes forwarding the response to a supervisor device when the required steps are absent from the performed steps of the response process.