Agent Interface for Automated Call Center Intervention
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Automated call center systems often frustrate callers when they cannot resolve issues, leading to increased wait times and agent workload due to the need for live agent intervention after automated systems fail, and existing data collection methods are inefficient and negatively impact user experience.
Innovation Solution
An agent interface system that allows live agents to access and intervene in automated call center interactions by monitoring conversation flow, semantic information, and recognized utterances with confidence levels, enabling timely intervention and reducing repetition, while also providing a user interface for agents to manage dialog systems effectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If callers are transferred to live agents only after automated system failure, then agent intervention is necessary to resolve issues, but caller frustration increases and agent workload increases due to repeated information gathering
Solution Approach 1:
The system performs preliminary actions by collecting caller information and context data during the automated interaction phase, before the caller is transferred to a live agent. This preliminary data gathering includes capturing speech transcripts, identified intents, and interaction history, which are then made available to the agent through the interface system, eliminating the need for callers to repeat information and reducing frustration.
Solution Approach 2:
The patent introduces an intermediary system - the agent interface - that mediates between the automated dialog system and the live agent. This interface captures and presents relevant context information from the automated interaction to the agent, serving as a bridge that transfers necessary data without requiring direct caller repetition, thus reducing caller frustration while maintaining reliable issue resolution.
2Productivity
If automated dialog systems handle all initial interactions, then agent productivity increases by reducing routine tasks, but data collection for system improvement becomes inefficient and requires negative user experiences
Solution Approach 1:
The system enables self-service data collection by automatically capturing and storing interaction data, speech transcripts, and context information during normal automated dialog operations. The agent interface automatically logs this data without requiring separate data collection sessions or user participation beyond the normal interaction, thus maintaining high agent productivity while efficiently collecting training data.
Solution Approach 2:
The patent implements continuous data collection during normal automated dialog operations rather than requiring separate data collection phases. The agent interface continuously captures interaction data, speech patterns, and context information as part of the ongoing service delivery, ensuring that data collection is an continuous useful action that occurs alongside productivity-enhancing automated interactions.
3Object-affected harmful factors
If live agents monitor all automated dialog sessions, then early intervention is possible to reduce frustration, but system complexity and operational costs increase
Solution Approach 1:
The system applies local quality by enabling agent monitoring and intervention only for specific dialog sessions where it is most beneficial, rather than requiring all sessions to be monitored. The agent interface allows selective access to automated dialog sessions based on criteria such as detected frustration indicators, complex intent recognition failures, or caller requests, thus reducing system complexity while still providing early intervention where needed to reduce caller frustration.
Data Source
AI summary
Embodiments of an interface system that enables a call center agent to access and intervene in an interaction between an automated call center system and a caller whenever necessary for complex application tasks is described. The system includes a user interface that presents the agent with one or more categories of information, including the conversation flow, obtained semantic information, the recognized utterances, and access to the utterance waveforms. This information is cross-linked and attached with a confidence level for better access and navigation within the dialog system for the generation of appropriate responses to the caller.