Call Control System with Real-Time Speech Playback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current call center systems do not allow supervisors to directly hear actual calls between customers and operators, limiting their ability to monitor and address potential issues in real-time, despite providing text-based content analysis.
Innovation Solution
A call control system that includes a call recorder, speech recognizer, recognition result manager, and information processor, enabling real-time speech data recognition, keyword detection, and playback of calls, allowing supervisors to directly hear and monitor conversations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a supervisor only confirms call content through text display, then the system complexity is reduced, but the monitoring effectiveness and ability to directly hear actual calls is insufficient
Solution Approach 1:
The patent combines text-based call content display with audio playback functionality into a single integrated system. The information processor simultaneously handles both text recognition results and audio signal transmission, allowing supervisors to monitor calls through both visual and auditory channels without requiring separate systems.
Solution Approach 2:
The information processor is designed to perform multiple functions: it displays recognized text content on the display unit, transmits audio signals to headphones, and coordinates keyword-based filtering. This multi-functional design enables a single device to provide comprehensive monitoring capabilities without increasing overall system complexity.
2Reliability
If the supervisor monitors all calls in real-time through audio playback, then the monitoring completeness is improved, but the monitoring operation efficiency decreases due to increased workload
Solution Approach 1:
Instead of requiring supervisors to monitor all calls completely, the system performs partial monitoring by automatically displaying only calls containing predetermined keywords. This selective approach allows supervisors to focus on potentially problematic calls while maintaining overall monitoring completeness for quality assurance.
Solution Approach 2:
The speech recognizer performs preliminary analysis of call content before the supervisor reviews it. By pre-identifying keywords and filtering calls automatically, the system prepares relevant information in advance, reducing the supervisor's workload and enabling faster response to important calls.
3Loss of information
If the system provides detailed text display of call content, then the information availability is improved, but the real-time audio monitoring capability is lost
Solution Approach 1:
The system merges text display and audio playback functions within the same information processor, allowing supervisors to simultaneously view transcribed call content and listen to actual audio. This integration ensures that neither text information nor audio monitoring capability is lost, but both are available together for comprehensive call analysis.
Data Source
AI summary
An information processor requests a recognition result manager to transmit recording information about a call including a keyword and a recognition result of speech recognition using an extension number as a key. The manager transmits the recording information about the call including the keyword corresponding to the extension number and the recognition result of the speech recognition to the processor. The processor displays a recognition result of speech recognition of the call including the keyword on a display unit. Upon receiving an input of an instruction to perform speech playback, the processor transmits recording information in association with text displayed on the display unit to a recorder. The recorder transmits speech data corresponding to the recording information to the processor. The processor plays back speech data corresponding to the recording information.