Generative AI Digital Assistant Execution Plan
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional digital assistants rely on predefined skill or intent models, limiting their ability to engage in intelligent and contextual conversations, as they require specific keywords or commands, hindering natural communication and understanding of user queries.
Innovation Solution
Implementing a computer-implemented method using generative artificial intelligence that constructs an input prompt from natural language utterances, generates an execution plan by determining suitable agents and actions, and executes the plan to provide contextually relevant responses, leveraging large language models (LLMs) and external knowledge sources like knowledge graphs and APIs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If traditional predefined skill or intent models are used, then the digital assistant can operate with simple keyword-based commands, but the ability to engage in intelligent and contextual conversations is limited
Solution Approach 1:
The patent transforms the operational parameters of the digital assistant from fixed keyword matching to dynamic natural language processing. By changing the underlying model from predefined intents to generative AI with LLMs, the system maintains operational simplicity while dramatically improving conversational adaptability and contextual understanding
Solution Approach 2:
The patent replaces the mechanical keyword-matching system with a neural network-based natural language processing system. This substitution enables the digital assistant to understand and generate human-like language while maintaining the simplicity of interaction, resolving the contradiction between operational simplicity and conversational intelligence
2Measurement precision
If predefined keywords or commands are required, then the digital assistant can process queries with clear structure, but natural communication and understanding of user queries are hindered
Solution Approach 1:
The patent introduces dynamic natural language processing capabilities that adapt to various user input styles and contexts. The system can dynamically adjust its understanding and response generation based on the conversational context, maintaining query structure clarity while enabling natural communication without requiring predefined keywords
3Ease of operation
If live communication channels with human service personnel are maintained, then real-time conversations can be conducted, but costs for organizations increase
Solution Approach 1:
The patent implements a self-service digital assistant using generative AI that can independently handle real-time conversations without human intervention. The system autonomously processes user queries, generates contextually appropriate responses, and manages conversations, thereby maintaining real-time interaction capability while eliminating the need for human service personnel and reducing operational costs
Data Source
AI summary
Techniques are disclosed herein for implementing digital assistants using generative artificial intelligence. An input prompt comprising a natural language utterance and candidate agents and associated actions can be constructed. An execution plan can be generated using a first generative artificial model based on the input prompt. The execution plan can be executed to perform actions included in the execution plan using agents indicated by the execution plan. A response to the natural language utterance can be generated by a second generative artificial intelligence model using one or more outputs from executing the execution plan.