Dynamic Service Dispatching Using Agent Skill Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing service dispatching systems in non-manufacturing domains like IT and healthcare lack the ability to dynamically assess service request complexity, agent skill variability, and non-stationary variance, leading to inefficient allocation of requests and suboptimal system performance.
Innovation Solution
A method that involves obtaining attributes of service requests and agents, incorporating feedback from agent queues, and using these attributes to determine suitable agents through algorithms that consider service complexity, priority, and agent skill variability, ensuring timely and cost-effective request fulfillment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If existing ad hoc methods are used to determine service request complexity and match with agent skills, then the dispatching process is simple to implement, but the service level and matching accuracy deteriorate
Solution Approach 1:
The patent transforms the dispatching system from using static, ad hoc complexity determination to using dynamic parameters including service request complexity scores, agent skill levels, and skill variability measures. These parameters are continuously updated based on feedback from queue states and performance metrics, enabling precise matching while maintaining systematic implementation through automated calculation.
Solution Approach 2:
The system incorporates feedback loops where queue states, service outcomes, and agent performance are continuously monitored and fed back into the dispatching algorithm. This feedback mechanism allows the system to learn from past dispatch decisions and improve matching accuracy over time without requiring manual reconfiguration.
2Ease of operation
If existing dispatching control systems assume stationary variance with constant agent skills, then the system is simple to manage, but the adaptability to changing service environments deteriorates
Solution Approach 1:
The patent implements a dynamic dispatching system where agent skill levels and skill variability are not fixed but change over time based on training, experience, and current queue states. The system continuously updates these parameters to reflect the current service environment, enabling adaptation to changing conditions while maintaining automated management through algorithmic updates.
Solution Approach 2:
The system changes the parameters from stationary assumptions to dynamic variables. Agent skills are represented as time-varying parameters that incorporate training effects and experience accumulation. The dispatching algorithm adjusts to these changing parameters automatically, providing adaptability without manual intervention.
3Speed
If existing approaches do not consider overall system state, then the dispatching decision process is fast and simple, but the overall system performance and service levels deteriorate
Solution Approach 1:
The system performs preliminary actions by pre-calculating and maintaining up-to-date parameters for service request complexity, agent skills, and queue states. This preparation allows the actual dispatching decision to be made quickly by comparing current request attributes against pre-computed optimal matches, achieving both speed and performance.
Solution Approach 2:
The system uses feedback from overall queue states to inform dispatching decisions. Queue length, wait times, and agent utilization metrics are continuously monitored and fed back into the dispatching algorithm, enabling system-wide optimization while maintaining fast decision-making through automated real-time adjustments.
Data Source
AI summary
Techniques for dispatching one or more services requests to one or more agents are provided. The techniques include obtaining one or more attributes of each service request, obtaining one or more attributes of each agent, obtaining feedback from each of one or more agent queues, and using the one or more attributes of each service request, the one or more attributes of each agent and the feedback from each of the one or more agent queues to determine one or more suitable agents to receive a dispatch for each of the one or more service requests. Techniques are also provided for generating a database of one or more attributes of one or more service requests and one or more attributes of one or more agents.