Dynamic Call-Progress Analysis for IVR Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional call progress analysis systems face timing issues in identifying whether a receiver is human or a machine, leading to delayed interactions or inappropriate voice applications, as they often take too long to distinguish between different types of receivers, causing user impatience or incorrect message delivery.
Innovation Solution
The call stream is simultaneously fed to both the call-progress analyzer and the IVR, allowing the IVR to dynamically adapt to real-time analysis results, enabling immediate interaction and potential session adjustments or replacements based on updated analysis, thereby minimizing delays and improving interaction accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional call progress analysis is used to identify receiver type, then measurement precision is improved, but loss of time increases
Solution Approach 1:
The system performs preliminary analysis of call progress indicators (such as detecting answering machine beeps or voice mail prompts) during the initial call setup phase. By analyzing these early indicators before the full interaction begins, the system pre-determines the receiver type (human, answering machine, or voice mail), allowing the IVR to adapt its behavior immediately without waiting for complete call establishment. This preliminary detection resolves the contradiction by providing accurate identification early in the call lifecycle.
Solution Approach 2:
The system continuously monitors call progress indicators and uses this feedback to dynamically adjust the IVR interaction. The call progress analysis module provides real-time feedback about the receiver's responses (such as detecting when an answering machine plays its greeting or when a human user engages with the IVR), allowing the system to adapt its behavior accordingly. This feedback mechanism enables accurate receiver identification while minimizing call setup time by responding to indicators as they occur.
2Measurement precision
If call progress analysis waits for signature beep to identify answering machine, then measurement precision is improved, but loss of time increases
Solution Approach 1:
The system performs preliminary detection of answering machine indicators during the initial call setup phase, analyzing call progress indicators such as detecting the characteristic beep or greeting message before the full interaction begins. By pre-detecting these indicators early in the call lifecycle, the system determines that the receiver is an answering machine without waiting for the complete signature beep sequence, thus resolving the contradiction between accurate detection and time efficiency.
Solution Approach 2:
The system analyzes partial call progress indicators (such as the initial portion of an answering machine's greeting or the first detectable beep) rather than waiting for the complete signature sequence. This partial analysis provides sufficient information to identify the receiver type as an answering machine, allowing the IVR to adapt its behavior without waiting for the full beep sequence, thereby reducing wait time while maintaining adequate detection accuracy.
3Productivity
If IVR interacts promptly assuming human receiver, then productivity is improved, but reliability worsens
Solution Approach 1:
The system dynamically adapts the IVR interaction based on real-time call progress analysis. Rather than statically assuming the receiver is human, the system continuously monitors call progress indicators (such as detecting answering machine beeps, voice mail prompts, or human user engagement patterns) and adjusts the IVR behavior accordingly. This dynamic adaptation allows the system to maintain high productivity by quickly responding to human users while ensuring reliability by switching to appropriate protocols when machines are detected.
Solution Approach 2:
The system uses real-time feedback from call progress analysis to adjust IVR interaction appropriateness. The call progress analysis module continuously provides feedback about the receiver's responses, allowing the IVR to adapt its behavior in real-time. This feedback mechanism ensures that the system maintains high productivity by promptly interacting with human users while ensuring reliability by detecting and adapting to machine receivers, thus resolving the contradiction between speed and appropriateness.
Data Source
AI summary
A telephony application such as an interactive voice response (“IVR”) needs to identify quickly the nature of the call (e.g., whether it is a person or machine answering a call) in order to initiate an appropriate voice application. Conventionally, the call stream is sent to a call-progress analyzer (“CPA”) for analysis. Once a result is reached, the call stream is redirected to a call processing unit running the IVR according to the analyzed result. The present scheme feeds the call stream simultaneous to both the CPA and the IVR. The CPA is allowed to continue analyzing and outputting a series of analysis results until a predetermined result appears. In the meantime, the IVR can dynamically adapt itself to the latest analysis results and interact with the call with a minimum of delay.