ERP Generative AI Query Responses With Database-Aware Prompting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing generative AI models integrated into ERP systems face challenges due to their pretraining on public data, lacking understanding of specific database information, leading to misaligned outputs and potential data integrity issues when making changes to database tables.
Innovation Solution
The integration of prompt engineering within the ERP system to dynamically retrieve data from database tables and metadata, embedding this information into prompt templates to guide the generative AI in generating contextually relevant responses, while configuring prompts to prevent unintended user actions and ensure data policy alignment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If existing generative AI models pretrained on public data are integrated into ERP systems, then automation and content creation capabilities are enhanced, but the AI models lack understanding of specific database information leading to misaligned outputs
Solution Approach 1:
The patent introduces an intermediary layer between the pre-trained generative AI model and the ERP database. This intermediary retrieves specific database information and metadata, then embeds it into prompt templates that guide the AI model. This mediator ensures the AI has access to organization-specific contextual information while maintaining the benefits of pre-trained automation capabilities.
Solution Approach 2:
The system performs preliminary actions by retrieving and embedding relevant database information and metadata into prompt templates before the AI model generates responses. This pre-preparation ensures the AI model receives contextually appropriate information in advance, improving alignment with organization-specific data without requiring retraining.
2Productivity
If generative AI models are allowed to make changes to database tables, then decision-making and data processing are enhanced, but data integrity and security risks increase
Solution Approach 1:
The patent implements feedback mechanisms where the system monitors AI-generated changes to database tables. By configuring prompts to prevent unintended user actions and by having the AI retrieve and validate information against existing database constraints and metadata, the system ensures that productivity gains do not compromise data integrity or security.
3Loss of information
If prompt engineering is used to dynamically retrieve and embed database information, then contextual relevance of AI responses is improved, but system complexity increases
Solution Approach 1:
The patent creates a universal prompt template system that can dynamically adapt to different database tables and information types. Rather than creating separate complex systems for each data type, the same prompt engineering framework handles various contexts by retrieving appropriate metadata and information, reducing overall system complexity while maintaining high contextual relevance.
Data Source
AI summary
A computer-implemented method for intelligent query response in an enterprise resource planning (ERP) system is disclosed. The method can receive, from a user interface of the ERP system, a natural language user query for processing data of a target object and invoke a function call through an application programming interface of the ERP system based on the user query. Invoking the function call includes extract denormalized data from one or more database tables, generating a prompt using a prompt template and the denormalized data, prompting a large language model using the prompt, and receiving a response generated by the large language model. The method can generate an output on the user interface based on the response generated by the large language model. Related systems and software for implementing the method are also disclosed.