Orchestration Layer for Dynamic Agent-Matching Algorithm Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing contact center systems lack the ability to dynamically select the best agent-matching algorithms and mechanisms in real-time, failing to consider various vital contact center routing conditions and attributes for optimal customer-agent matching.
Innovation Solution
The implementation of an orchestration layer that selectively binds and configures multiple agent-matching algorithms, allowing for dynamic selection based on real-time contact center conditions, such as call surplus, agent proficiency, and queue metrics, to determine the best agent match through a cascading mechanism.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple agent-matching algorithms are bound together with an orchestration layer for dynamic selection, then the adaptability and accuracy of customer-agent matching is improved, but the device complexity and computational overhead increase
Solution Approach 1:
The system segments the agent-matching functionality into multiple independent algorithms (e.g., skill-based matching, performance-based matching, customer preference-based matching). Each algorithm operates as a separate module that can be independently configured, executed, and optimized. The orchestration layer acts as a coordinator that selects and combines these segmented algorithms based on real-time conditions, thereby achieving high adaptability without requiring a monolithic complex system.
Solution Approach 2:
The orchestration layer implements dynamic selection of agent-matching algorithms based on real-time contact center conditions such as call volume, agent availability, customer urgency, and algorithm performance metrics. The system can dynamically adjust which algorithms are active and how they are combined, allowing the matching process to adapt to changing conditions without manual reconfiguration, thus improving versatility while keeping the system structure manageable.
2Productivity
If dynamic configuration based on real-time routing conditions is implemented, then the productivity and efficiency of agent selection is improved, but the loss of time for configuration and selection increases
Solution Approach 1:
The system performs preliminary configuration of multiple agent-matching algorithms during system setup or off-peak periods. Algorithm parameters, weights, and selection criteria are pre-configured based on historical data and expected scenarios. When real-time routing decisions are needed, the orchestration layer simply selects from pre-configured options rather than performing complex configuration from scratch, thereby maintaining high productivity while minimizing real-time configuration time.
Solution Approach 2:
The orchestration layer continuously monitors the performance of each agent-matching algorithm and uses this feedback to dynamically adjust algorithm selection and parameters. Performance metrics such as match quality, agent satisfaction, and customer feedback are collected and fed back to the orchestration layer, which automatically adjusts the configuration to optimize productivity. This closed-loop feedback mechanism enables efficient adaptive configuration without requiring extensive manual intervention or time-consuming reconfiguration.
3Measurement precision
If cascading mechanism executes all available agent matcher modules, then the measurement precision of agent matching is improved, but the loss of time and computational resources increases
Solution Approach 1:
The cascading mechanism does not always execute all available agent-matching modules in full. Instead, the orchestration layer selectively activates only the necessary subset of algorithms based on the specific routing scenario, customer requirements, and current system state. For simple calls, only essential matching criteria are evaluated; for complex or urgent calls, additional algorithms are activated. This partial action approach maintains measurement precision for critical cases while reducing average execution time and resource consumption.
Solution Approach 2:
Different levels of matching precision are applied to different call scenarios and customer segments. The system identifies the specific requirements of each incoming call and activates only the matching algorithms relevant to that particular case. For example, routine inquiries may use basic skill-based matching, while complex technical issues trigger more sophisticated algorithms. This local quality approach ensures high precision where needed while minimizing unnecessary computational overhead for simpler cases.
Data Source
AI summary
Making the best customer-agent matching decision to engage in a communication to resolve a work item is a primary objective of any contact center. In addition to or as an alternative to standard routing decisions, a real-time agent-matching module is selected from a pool of agent-matching modules, each real-time agent-matching module being unique in its ability to match a resource (e.g., agent) to a customer. The real-time agent-matching modules may differ from one another in terms of information gathered and/or utilizing logic unique to that particular agent-matching module. An orchestration layer is provided to select a single agent-matching module or a series or parallel number of agent-matching modules to select a resource and connect to the customer.