Method and system for matching entities in an auction

What is this patent about?

'086 is related to the field of intelligent communication routing and automated resource allocation within computer-telephony integrated (CTI) systems. It addresses the technical challenge of efficiently matching entities, such as callers and agents in a call center, by moving beyond simple rule-based or first-in-first-out (FIFO) logic. The background context involves managing high-volume real-time communications where the goal is to balance service quality with resource utilization, particularly in environments where agents possess varying skill levels and costs.

The underlying idea behind '086 is to treat the matching of entities as a multivariate optimization problem that accounts for both immediate economic gain and the long-term cost of resource unavailability. Instead of just looking for the best available agent for a specific caller, the system evaluates the opportunity cost of using a specific resource now versus saving it for a potentially more valuable future interaction. This engineering insight allows the system to maximize the total economic surplus across a series of matches rather than just optimizing a single transaction in isolation.

The claims of '086 focus on a method for matching subsets of entities by processing multivalued scalar data that represent inferential targeting and characteristic parameters. The independent claims specifically protect the use of an automated processor to perform an optimization that balances the economic surplus of a mutually exclusive match against the opportunity cost of making those entities unavailable for alternate pairings. This mechanism ensures that the resulting signal for call routing or resource allocation reflects a mathematically optimized selection across a plurality of potential matches.

In practice, the invention functions by maintaining a local database of agent skill vectors and caller requirement vectors within the low-level switching architecture. By integrating the optimization algorithm directly into the communication server rather than externalizing it to a high-level management system, the '086 patent reduces processing latency and communication overhead. The system can dynamically adjust to peak loads by shifting from complex training-focused optimizations to high-throughput efficiency modes, ensuring that the most critical matches are prioritized when resources are scarce.

This approach differs from prior solutions by incorporating game theory and economic modeling into the real-time switching logic. Traditional systems typically use static grouping or simple skill-based thresholds that ignore the relative value of an agent's time or the potential for skill growth. By quantifying abstract concepts like training value and future availability into a normalized cost function, the '086 patent allows for a more granular and globally efficient distribution of work that adapts to the shifting state of the entire call center ecosystem.