Chatbot Intent Routing to Balance AI Relevance and Accuracy

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing chatbot systems struggle to provide accurate and contextually relevant responses due to the limitations of gen-AI chatbots, which can generate unpredictable and incorrect information, while non-gen-AI chatbots may fail to provide contextually relevant responses for certain user inputs, leading to suboptimal user experiences and inefficient resource usage.

Innovation Solution

A chatbot risk management system that determines user intent and selectively chooses between gen-AI and non-gen-AI chatbot services based on intent information, using intent blocklists and allowlists to ensure accurate and relevant responses, thereby optimizing resource usage.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If gen-AI chatbot service is used to generate responses, then contextual relevance and conversational flexibility are improved, but response accuracy and reliability deteriorate due to unpredictable hallucinations

Engineering Contradiction:

Improvecontextual relevanceVSAvoidresponse accuracy

Core Design Contradiction:

Adaptability or versatilityVSReliability

Solution Approach 1:

The patent segments the chatbot response generation process into two distinct paths: gen-AI chatbot service for contextual relevance and non-gen-AI chatbot service for accuracy. The system divides user inputs into different categories and routes them to appropriate services, preventing gen-AI hallucinations from affecting all responses.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces an intermediary classification mechanism that sits between user input and chatbot services. This intermediary determines which service to use based on input characteristics, acting as a mediator that selects the appropriate response generation path to balance accuracy and contextual relevance.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Reliability

If non-gen-AI chatbot service is used to ensure accurate responses, then response accuracy is improved, but contextual relevance and conversational flexibility worsen

Engineering Contradiction:

Improveresponse accuracyVSAvoidcontextual relevance

Core Design Contradiction:

ReliabilityVSAdaptability or versatility

Solution Approach 1:

The system segments chatbot services into specialized components: non-gen-AI service handles accuracy-critical inputs while gen-AI service handles contextual-relevance-critical inputs. This segmentation allows each service to excel at its designated function without compromising overall system performance.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent implements dynamic service selection where the system adapts its response generation approach based on the characteristics of each user input. The classification mechanism dynamically determines whether to use gen-AI or non-gen-AI service, allowing the system to be flexible and context-aware.

Inventive Principle:

Principle #15Dynamics

3Adaptability or versatility

If gen-AI chatbot service is used for all user inputs, then contextual relevance is improved, but resource consumption increases

Engineering Contradiction:

Improvecontextual relevanceVSAvoidcomputing resource consumption

Core Design Contradiction:

Adaptability or versatilityVSUse of energy by moving object

Solution Approach 1:

The patent applies partial action by using gen-AI chatbot service only for inputs that require its contextual understanding capabilities, rather than applying it to all inputs. This selective approach reduces unnecessary resource consumption while maintaining contextual relevance where needed.

Inventive Principle:

Principle #16Partial or excessive action

Solution Approach 2:

The system changes the operational parameters of chatbot service selection based on input characteristics. By adjusting which service is activated based on input classification, the system optimizes resource usage while maintaining appropriate response quality for different types of queries.

Inventive Principle:

Principle #35Parameter changes

4Reliability

If multiple chatbot services are maintained for different input types, then response quality is improved, but system complexity increases

Engineering Contradiction:

Improveresponse qualityVSAvoidsystem architecture

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent implements a universal classification mechanism that handles multiple input types and routes them to appropriate services. This multi-functional classifier manages the complexity of multiple chatbot services by providing a unified interface for service selection.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The classification system acts as an intermediary layer that manages the complexity of multiple chatbot services. This mediator handles the routing logic and service coordination, shielding the user interface from the underlying system complexity while enabling response quality improvement.

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS20260030363A1Chatbot risk management

Publication Date: 2026.01.29 CAPITAL ONE SERVICES LLC

AI summary

Some implementations described herein relate to a system for chatbot risk management. The system is configured to receive, from a user device that includes a chatbot interface, user input associated with the chatbot interface. The system is configured to determine, based on the user input, intent information. The system is configured to select, based on the intent information, a chatbot service from a generative-artificial-intelligence (gen-AI) chatbot service and a non-gen-AI chatbot service. The system is configured to provide the user input to the selected chatbot service to allow the selected chatbot service to generate a response to the user input.