Generative AI Smart Answering System with Curated Knowledgebase

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional automated customer service systems face challenges due to rigid data constraints, leading to ineffective handling of customer queries and a frustrating customer experience, while AI systems often fabricate information, providing false responses.

Innovation Solution

A smart answering system utilizing generative artificial intelligence to build and utilize a knowledgebase, where labels are generated by scraping merchant websites and edited by merchants, ensuring accurate responses by limiting AI responses to the information within the knowledgebase.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If conventional automated customer service systems use rigid data constraints, then system structure is simplified, but the ability to handle a wide range of customer queries deteriorates

Engineering Contradiction:

Improvesystem structureVSAvoidability to handle customer queries

Core Design Contradiction:

Device complexityVSAdaptability or versatility

Solution Approach 1:

The system transforms structured data into unstructured natural language responses dynamically. Instead of using rigid predefined constraints, the system generates responses by combining relevant information fragments in natural language, allowing flexible adaptation to various customer queries while maintaining manageable system structure through modular information organization.

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The system employs dynamic response generation where the structure and content of answers are determined at runtime based on the specific customer query. The system can adapt its response format, detail level, and information selection dynamically, rather than following fixed predetermined paths, enabling handling of diverse queries with a unified system architecture.

Inventive Principle:

Principle #15Dynamics

2Adaptability or versatility

If AI systems generate responses freely, then response flexibility is improved, but information fabrication deteriorates

Engineering Contradiction:

Improveresponse flexibilityVSAvoidinformation accuracy

Core Design Contradiction:

Adaptability or versatilityVSReliability

Solution Approach 1:

The system introduces a knowledge base as an intermediary between the customer query and the AI response generator. This knowledge base contains verified information fragments that constrain and guide the AI's response generation, ensuring that flexible natural language responses are grounded in accurate, pre-verified information rather than AI hallucinations.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The system implements feedback mechanisms where customer queries are analyzed to identify relevant information needs, which then guide the selection and combination of appropriate knowledge base entries. The system continuously refines its responses based on query analysis feedback, ensuring both flexibility in response formulation and accuracy through verification against stored knowledge.

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS20250029114A1Artificial intelligence smart answering architecture

Publication Date: 2025.01.23 SOUNDHOUND AI IP LLC

AI summary

An automated answering system and method are disclosed for use in providing automated customer service. The automated answering system uses generative artificial intelligence to aid in forming a knowledgebase of information regarding a merchant's business that is used in answering the customer queries. The automated answering system of the present technology also uses generative artificial intelligence to aid in formulating a response to queries using the formed knowledgebase.