Tunable Chatbot Personality Control in Contact Centers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Contact centers face challenges in dynamically adjusting automated agents or chatbots to optimize customer interactions, as existing systems lack real-time monitoring and modification capabilities, and struggle to select the most effective chatbot based on situational context and customer preferences.
Innovation Solution
Implementing a system that allows contact centers to monitor customer interactions with automated agents or chatbots in real-time, adjust their characteristics, and pre-test chatbots before customer interactions, using AI modeling to select the best chatbot based on predicted outcomes and cost-benefit analysis, while considering different AI-agent providers and their training data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple chatbots from different AI-agent providers are deployed in a contact center, then the versatility and adaptability of the system improves, but the device complexity and difficulty of selecting the most effective chatbot increases
Solution Approach 1:
The patent introduces an intermediary system that sits between multiple AI-agent providers and the contact center operations. This intermediary abstracts the complexity of managing multiple chatbots by providing a unified interface for deployment, monitoring, and selection, thereby maintaining versatility while reducing system complexity
Solution Approach 2:
The system creates a universal platform that can handle multiple chatbots from different providers through a common set of controls and dials. This multi-functional approach allows the same infrastructure to manage diverse chatbot types, simplifying the overall system architecture
2Reliability
If real-time monitoring and modification of chatbot characteristics is implemented, then the customer interaction quality improves, but the use of energy and computational resources increases
Solution Approach 1:
The system implements periodic monitoring and modification of chatbot characteristics rather than continuous real-time adjustments. This allows the system to maintain interaction quality by checking and adjusting chatbot parameters at appropriate intervals, reducing unnecessary computational overhead while still responding to significant changes in interaction patterns
Solution Approach 2:
The patent employs feedback mechanisms where chatbot performance is monitored and characteristics are modified based on observed outcomes. This feedback loop ensures interaction quality improves by learning from actual performance data, while computational resources are used efficiently by only making adjustments when feedback indicates improvement is needed
3Reliability
If pre-testing of chatbots is performed before customer interactions, then the reliability of customer service improves, but the loss of time in the interaction process increases
Solution Approach 1:
The system performs preliminary testing of chatbots during off-peak periods or in parallel with other operations, so that by the time customer interactions need to be routed, the testing and selection process is already complete. This preliminary action ensures service reliability without adding time delays to actual customer service operations
Solution Approach 2:
The patent implements dynamic chatbot selection where the system adapts its testing and selection process based on current conditions. For routine interactions, pre-tested chatbots are used immediately, while for novel or complex situations, the system dynamically selects which chatbots to test, balancing reliability requirements with time constraints
Data Source
AI summary
A contact center, communication system, and method of operating a contact center are provided. In one example, the contact center includes a tunable chatbot stored in memory and a chatbot tuning function that controls and adjusts a personality characteristic of the tunable chatbot while the tunable chatbot is interacting with a customer via a customer communication device.