Agent Contact Matching via Weighted Bipartite Graph
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing call center systems fail to match contacts with agents based on attributes, leading to poor interaction metrics such as long handle times, low customer satisfaction, and poor sales conversion rates, as they do not consider the attributes of agents and contacts during the matching process.
Innovation Solution
A contact center service uses a machine learning model to generate a weighted bipartite graph, where edge weightings represent predicted interaction metrics between agents and contacts, allowing for intelligent matching based on attributes like time zones, language skills, and demographics, and applies matching rules to determine optimal agent-contact pairs for improved interaction outcomes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If first-come, first-serve matching is used, then queue management is simple, but interaction metrics deteriorate (long handle times, low customer satisfaction)
Solution Approach 1:
The system changes the matching parameter from simple queue position to attribute-based compatibility scores. The machine learning model evaluates multiple parameters including language skills, time zone, demographic attributes, and agent expertise to determine optimal matches, transforming the matching criterion from temporal (queue order) to qualitative (attribute compatibility).
Solution Approach 2:
The patent introduces an intermediary matching service that sits between the queue and agents. This intermediary evaluates attribute compatibility and selects optimal agent-contact pairs, mediating the connection process to improve interaction metrics while maintaining queue management functionality.
2Reliability
If attribute-based matching is implemented, then interaction metrics improve, but system complexity increases
Solution Approach 1:
The system implements self-service through automated machine learning models that independently evaluate attribute compatibility and make matching decisions without requiring complex manual configuration or intervention. The model automatically processes agent and contact attributes to generate optimal matches.
Solution Approach 2:
The patent applies preliminary action by pre-processing and storing agent and contact attributes in structured formats before matching occurs. The machine learning model is pre-trained on historical interaction data, enabling it to quickly evaluate compatibility without complex real-time computations during actual matching.
3Measurement precision
If machine learning model is used for matching, then matching accuracy improves, but computational requirements increase
Solution Approach 1:
The system applies partial action by evaluating only the most relevant attributes for each matching scenario rather than processing all possible attributes equally. The machine learning model identifies and weights key attributes based on their importance to interaction success, reducing computational overhead while maintaining accuracy.
Data Source
AI summary
Technology is described for routing communications between contacts and agents. Associations may be generated between agents and contacts who intend to communicate with the agents. Weightings for the associations between the agents and the contacts may be determined using a machine learning model to produce weighted associations. The weightings may represent a predicted interaction metric between the agents and the contacts. Selected pairs of agents and contacts may be determined by applying matching rules to the weightings. Communications may be routed between the contacts and the agents in accordance with the selected pairs of agents and contacts.