Real-Time Interaction Analysis System for Call Center Alerting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing call center systems lack real-time or near-real-time alert mechanisms to identify interactions requiring supervisor intervention, leading to potential customer churn, unsatisfied customers, or fraud detection issues.
Innovation Solution
A method and apparatus that classify interactions in real-time by extracting features such as emotional scores, talk analysis parameters, and global features, applying a model to determine if an interaction is associated with problems like customer dissatisfaction or fraud, and taking actions like generating alerts or routing calls.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If real-time classification of interactions is implemented, then supervisor intervention can be timely provided, but system complexity increases
Solution Approach 1:
The interaction is divided into segments that are processed independently in real-time. Each segment is classified separately using the trained model, allowing the system to handle complex analysis through manageable chunks without overwhelming system resources.
Solution Approach 2:
A classification model is trained in advance using a training corpus of interactions. This preliminary training phase prepares the model to perform real-time classification without requiring complex computations during actual interaction processing, thus reducing runtime system complexity.
2Measurement precision
If multiple features are extracted and aggregated in real-time, then classification accuracy improves, but processing time increases
Solution Approach 1:
Feature extraction is performed on segmented portions of interactions rather than entire interactions at once. This allows multiple features to be extracted and aggregated incrementally, maintaining classification accuracy while reducing the computational burden at any given moment.
Solution Approach 2:
The system extracts and aggregates features from segments as they become available, rather than waiting for complete interactions. This partial action approach provides timely classifications with sufficient accuracy without requiring full processing of all possible features.
3Reliability
If real-time alert generation is implemented, then customer churn can be reduced, but operational resources increase
Solution Approach 1:
The system automatically generates alerts and notifications based on real-time classification results without requiring constant human monitoring. The automated alert generation system serves itself by processing interactions, classifying them, and notifying supervisors only when intervention is needed, thus maintaining customer retention while preserving operational efficiency.
Data Source
AI summary
A method and system for indicating in real time that an interaction is associated with a problem or issue, comprising: receiving a segment of an interaction in which a representative of the organization participates; extracting a feature from the segment; extracting a global feature associated with the interaction; aggregating the feature and the global feature; and classifying the segment or the interaction in association with the problem or issue by applying a model to the feature and the global feature. The method and system may also use features extracted from earlier segments within the interaction. The method and system can also evaluate the model based on features extracted from training interactions and manual tagging assigned to the interactions or segments thereof.