Call Routing via N-Gram Matching and Predictive Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional call center routing methods often fail to efficiently identify high business value callers in real-time, leading to long wait times and dissatisfaction, as they rely on 'round-robin' techniques that do not consider caller identity or value, resulting in inadequate allocation of resources.
Innovation Solution
The implementation of a predictive machine learning model that uses n-gram modeling and approximate string matching analysis to identify and prioritize callers based on customer identifiers such as name, address, and phone number, routing them to appropriate agents or groups within the call center.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional round-robin routing methods are used, then all callers are treated equally with simple routing logic, but high business value callers experience long wait times and dissatisfaction
Solution Approach 1:
The patent implements value-based routing that treats different callers differently based on their identified business value. High-value callers are routed to priority queues with dedicated agents, while lower-value callers follow standard routing procedures. This local differentiation in service quality resolves the contradiction by optimizing wait times for high-value callers without completely abandoning equitable treatment for all callers.
Solution Approach 2:
The system changes the routing parameter from simple sequential assignment (round-robin) to multi-dimensional criteria including caller value assessment, agent availability, and queue status. By introducing value scoring and dynamic parameter adjustment, the system can prioritize high-value callers while maintaining overall routing efficiency through automated decision-making.
2Reliability
If real-time caller identification and value assessment is implemented, then high business value callers can be prioritized, but system complexity increases due to data processing requirements
Solution Approach 1:
The system performs preliminary caller identification and value assessment actions during the call setup phase by analyzing caller ID, CRM data, and historical interaction records before the caller reaches the queue. This preliminary processing enables accurate value classification without adding complexity to the real-time routing decision, as the assessment is completed in advance.
Solution Approach 2:
The patent introduces an intermediary value assessment module that sits between the call reception system and the routing logic. This intermediary component handles the complex data processing, caller identification, and value scoring, while presenting simplified routing decisions to the core switching system. This mediator architecture isolates complexity to a dedicated component rather than distributing it throughout the entire routing system.
3Productivity
If automated call routing with value assessment is implemented, then resource allocation to high-value callers improves, but errors in customer identification can occur
Solution Approach 1:
The system implements beforehand cushioning by incorporating error tolerance and validation mechanisms in the caller identification process. Multiple data sources (caller ID, CRM records, historical data) are cross-referenced to verify caller identity before assigning value scores. This preliminary validation cushions against identification errors that could lead to incorrect routing decisions.
Solution Approach 2:
The patent implements feedback loops where routing outcomes and caller responses are continuously monitored and fed back into the value assessment model. When identification errors are detected (such as callers being routed to wrong agent groups or expressing dissatisfaction), the system learns from these feedback signals and adjusts its identification algorithms to improve future accuracy while maintaining efficient resource allocation.
Data Source
AI summary
A call management system of a call center identifies an inbound caller based upon computer analysis of customer identifiers, which may include at least two of customer name, street address, and zip code. Approximate string matching analysis matches n-grams generated from strings within customer identifiers, with n-grams generated from customer identification fields while searching one or more databases. Approximate string matching can incorporate a closeness metric based on Jaccard distance, and a Gaussian mixture model of best matches. In one embodiment, a polymr search engine analyzes customer identifiers of inbound callers to retrieve customer data, such as customer demographic data, matched to the customer identifiers. In another embodiment, the polymr search engine analyzes customer identifiers of inbound callers to identify repeat callers and retrieve previously collected customer data. Retrieved customer data is used in predictive modeling and scoring value of the inbound call, and in routing the scored inbound call.