Omni-channel Routing Broker for High-Throughput Service Centers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In multi-tenant environments, existing routing systems face challenges in efficiently routing service requests across a large pool of agents due to concurrent updates and distributed system complexities, leading to potential over-burdening of agents and bandwidth limitations, which hinders the ability to handle high volumes of incoming work effectively.
Innovation Solution
The omni-channel routing broker system uses a cascading series of queues and multithreading to manage requests, separating routing decisions from commitment, and employs a master agent presence and status data store with eventually consistent in-memory node-based databases to evaluate agent availability and make routing selections, ensuring efficient routing across multiple nodes and app servers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a distributed system with multiple app servers is used to handle large agent pools, then the system can serve more agents and handle more work, but concurrent updates and synchronization issues cause agents to be over-burdened and routing decisions to fail
Solution Approach 1:
The system segments the distributed architecture into multiple nodes, each maintaining its own local in-memory database for agent presence and status. This segmentation eliminates the need for concurrent access to a single shared database, preventing over-burdening while maintaining routing accuracy across the distributed system.
Solution Approach 2:
The patent introduces an intermediary mechanism where each node publishes routing decisions to an event queue rather than directly accessing shared resources. This intermediary approach allows asynchronous communication between nodes, ensuring reliable routing decisions without synchronization bottlenecks.
2Speed
If routing decisions are made in real-time across distributed nodes, then routing speed can be improved, but bandwidth consumption increases and synchronizing access to shared resources becomes challenging
Solution Approach 1:
The system performs preliminary actions by maintaining in-memory databases at each node that are proactively updated with agent presence and status information. This preliminary caching eliminates the need for real-time synchronization during routing decisions, reducing bandwidth consumption while maintaining fast routing speed.
Solution Approach 2:
Instead of continuous synchronization, the system uses periodic event-driven updates where nodes publish changes to the event queue at specific moments when agent status changes. This periodic action reduces bandwidth consumption compared to continuous real-time synchronization while maintaining up-to-date routing information.
3Device complexity
If a single app server is used to serve all agents in an organization, then routing decisions can be centralized and simplified, but the app server size becomes a limiting factor for scaling to large agent pools
Solution Approach 1:
The patent segments the centralized routing system into multiple distributed nodes, each capable of independently making routing decisions for portions of the agent pool. This segmentation removes the single app server bottleneck, enabling the system to scale to large agent pools while maintaining manageable complexity at each node through standardized in-memory database structures.
Data Source
AI summary
Omni-channel routing broker technology is usable to improve the experience for customers and for workers using service channels, especially for very large enterprise service operation centers that have large pools of agents, helping businesses determine the relative priority for handling a variety of service channels, and to efficiently route issues accordingly. Omni-channel routing broker includes intelligent routing of service requests in a large, distributed service center operation, efficiently prioritizing the routing of work across organizations to agents based on availability, capacity and priority—in a multi-tenant environment. The disclosed methods are usable for managing digital data for many tenants to software instances, including groups of users who share common access with a specific set of privileges to a software instance of at least one application. The disclosed technology makes possible the delivery of an improved performance from routing 3-4 requests per second to routing 100 requests per second.