Telephony control system with intelligent call routing

What is this patent about?

’979 is related to the field of computer-integrated telecommunications and intelligent switching architectures. Specifically, it addresses the management of call centers where incoming or outgoing communications must be efficiently paired with human agents. The background context involves the evolution from simple first-in-first-out queues to more complex skill-based routing systems that attempt to match the specific needs of a caller with the specialized expertise of an agent to maximize throughput and service quality.

The underlying idea behind ’979 is the integration of high-level intelligent decision-making directly into the low-level communications management architecture. Rather than relying on a separate, high-latency external database or management system to resolve a target address, the invention performs multivariate cost function optimizations within a common operating environment. This allows the system to evaluate complex, non-deterministic algorithms—such as those involving training benefits, opportunity costs, and agent proficiency—in real time to determine the absolute best pairing of a caller and an agent.

The claims of ’979 focus on a communications control system and method that utilize a common operating environment to both determine an optimum target and execute the routing. The independent claims specifically cover the use of a processor to evaluate a multivariate cost function comparing at least three potential targets based on call classification data and agent characteristics. Furthermore, the claims extend to a software architecture using a multithreaded operating system with dynamically linkable applications and a specific method for remediating discrimination by allocating skill-building opportunities through cost-benefit optimization.

In practice, the invention works by representing both calls and agents as vectors in a multidimensional skill space. When a call arrives, the system calculates a score for available agents by applying weights to these skill vectors. Unlike prior approaches that only look at immediate availability or static groups, this system can incorporate dynamic factors like training utility, where a call might be routed to a trainee to improve their skills, or a 'shadowing' mode where a trainer and trainee are paired on the same circuit. This ensures that the routing logic is not just a simple rule but a strategic economic calculation.

This approach differs from prior solutions by eliminating the rigid partitioning between the telephony switch and the intelligent policy manager. By offloading voice processing to dedicated peripherals while keeping the control logic on a high-speed general-purpose processor, the system avoids the latencies of external CRM lookups. It also moves beyond static 'primary/secondary' skill labels to a continuous adaptive profile that updates based on actual performance, allowing the call center to optimize for long-term organizational growth rather than just short-term queue reduction.