LLM Data Modeling from Catalog Assets with Less UI Navigation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current data modeling software tools require manual selection and configuration of data model elements, leading to inefficient user interfaces and increased complexity due to the need for navigating through multiple views and windows.

Innovation Solution

Implementing a generative data modeling system using a large language model to automatically generate and refine data models based on user input, reducing the need for manual navigation and configuration by utilizing a catalog of data assets, a metadata integration layer, and a generative modeler to process user inputs and generate data models.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If manual selection and configuration of data model elements is used, then users have full control over data model creation, but the user interface becomes inefficient and the process becomes complex due to navigating through multiple views and windows

Engineering Contradiction:
Improveease of data model creationVSAvoidinterface complexity
Core Design Contradiction:
Ease of operationVSDevice complexity

Solution Approach 1:

The system enables self-service data modeling by automatically generating data models from natural language descriptions. The generative modeler autonomously creates data model elements, relationships, and configurations without requiring manual intervention, thereby simplifying the user interface and improving ease of operation while maintaining full control over the data model creation process

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent replaces the mechanical interaction of manually navigating through multiple views and windows with a natural language-based interface. Users can describe their data modeling requirements in plain language, and the system automatically interprets and executes these requirements, eliminating the need for complex manual navigation and configuration

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Productivity

If manual configuration of every data model element is required, then precision and control are maintained, but productivity and efficiency decrease significantly

Engineering Contradiction:
Improvedata modeling efficiencyVSAvoiddata model configuration accuracy
Core Design Contradiction:
ProductivityVSManufacturing precision

Solution Approach 1:

The system incorporates feedback mechanisms where the generative modeler continuously refines data model elements based on user input and system-generated descriptions. The modeler generates natural language descriptions of data model elements, allows for user review and modification, and automatically adjusts configurations based on feedback, thereby maintaining high precision while significantly improving productivity

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The system performs preliminary actions by automatically generating data model elements, relationships, and configurations before user review. The generative modeler creates initial data models based on natural language descriptions, pre-configures elements according to best practices and data asset metadata, and prepares them for user validation, thereby reducing the time and effort required for manual configuration while maintaining accuracy

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentUS20250378064A1Generative data modeling using large language model
Publication Date: 2025.12.11 SAP SE
  • US20250378064A1 patent drawing
  • US20250378064A1 patent drawing
  • US20250378064A1 patent drawing

AI summary

A computer-implemented method may comprise receiving a first user input from a computing device and identifying a plurality of data assets from a catalog of data assets based on the first user input, where each data asset in the catalog of data assets comprises a corresponding entity that comprises data. The method may further comprise causing the identified plurality of data assets to be displayed on the computing device, receiving a first user selection of one or more of the identified plurality of data assets from the computing device, obtaining a plurality of data models based on the one or more of the identified plurality of data assets using a large language model, and causing the plurality of data models to be displayed on the computing device.