LLM-Generated Data Integrity Instructions for Diverse Data Schemas

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional online systems require significant manual effort and user interaction to create data integrity checks due to varying data formats, leading to increased complexity and time in data management.

Innovation Solution

An online system utilizes a generative model to automatically generate data integrity instructions by leveraging previously used instructions and metadata, reducing the need for manual configuration and syntax identification.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual creation of data integrity instructions is used, then data validation accuracy is improved, but time consumption and user effort increase significantly

Engineering Contradiction:

Improvedata validation accuracyVSAvoidtime consumption

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system performs preliminary action by automatically generating data integrity instructions before manual review, using generative AI to create validation rules, quality checks, and dashboard files based on schema information and sample data, thereby reducing the time required while maintaining accuracy through subsequent manual verification

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

A generative AI model serves as an intermediary between the raw data schema and the final data integrity instructions, automatically translating schema definitions into validation rules, quality checks, and dashboard configurations, thereby reducing manual effort while preserving validation accuracy

Inventive Principle:

Principle #24Intermediary (Mediator)

2Measurement precision

If manual identification of data portions and criteria specification is performed, then data integrity check accuracy is improved, but user interaction complexity increases

Engineering Contradiction:

Improvedata integrity check accuracyVSAvoiduser interaction complexity

Core Design Contradiction:

Measurement precisionVSEase of operation

Solution Approach 1:

The system enables self-service by automatically identifying data portions and specifying validation criteria using generative AI, which analyzes schema information and sample data to generate appropriate validation rules and quality checks without requiring users to manually identify each data portion or specify criteria, thereby reducing interaction complexity while maintaining accuracy

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The manual mechanical process of identifying data portions and specifying criteria is replaced by an automated AI-based system that generates data integrity instructions through natural language processing and generative models, thereby simplifying user interaction while preserving validation accuracy

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Adaptability or versatility

If data is stored in various formats for flexibility, then data storage versatility is improved, but data integrity check complexity increases

Engineering Contradiction:

Improvedata storage versatilityVSAvoiddata integrity check complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The system achieves universality by creating a unified data integrity instruction framework that can handle multiple data formats and schemas through generative AI, which automatically adapts validation rules and quality checks to different data structures, thereby maintaining storage versatility while reducing check complexity

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The system manages format diversity by dynamically adjusting validation parameters and criteria based on the specific data format being checked, using generative AI to generate format-appropriate validation rules automatically, thereby preserving storage versatility while simplifying the integrity check process

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS12487986B2Generating information integrity instructions using a generative model

Publication Date: 2025.12.02 MAPLEBEAR INC

AI summary

An online system stores data obtained from various users of the online system. For example, the online system maintains databases for various users, with a database including data received from the user. As users provide data to the online system for storage, the online system applies data integrity checks received from users that verifies received data satisfies one or more criteria. To facilitate creation and execution of data integrity checks, the online system tunes a large language model (LLM) using executable instructions for previously generated data integrity checks and metadata describing execution of the previously generated data integrity checks. After tuning, the online system obtains one or more parameters that are input as prompts to the LLM to generate executable instructions for performing a data integrity check using the parameters.