LLM-Graph-of-Thoughts for Unified GraphQL Schema Integration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems face challenges in handling frequent API changes, testing and integration efforts, security, and scalability in the context of Open Banking APIs, particularly in financial institutions, due to complex and dynamic data structures and security requirements.
Innovation Solution
A synergistic use of a Large Language Model (LLM) and Graph-of-Thoughts (GoT) is employed to provide unified data access across GraphQL APIs via a single API, enabling real-time query adjustments, automated API generation, and scalable data management through a schema hub and local traffic management.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple separate APIs are used to access different data sources, then data access coverage is improved, but system complexity and integration effort increase
Solution Approach 1:
The patent combines multiple separate GraphQL APIs into a single unified GraphQL API that can access data from multiple sources. The unified API consolidates what was previously multiple separate endpoints, allowing clients to query data from different sources through one interface, thereby reducing system complexity while maintaining comprehensive data access coverage.
Solution Approach 2:
The unified GraphQL API serves multiple functions by accessing diverse data sources through a single endpoint. It provides universal data access capabilities that work across different data sources and query types, eliminating the need for multiple specialized APIs while maintaining adaptability to various data access requirements.
2Reliability
If manual API modification is performed to handle frequent changes, then API functionality is maintained, but development time and costs increase
Solution Approach 1:
The system implements self-service capabilities through automated schema validation and query transformation. When API changes occur, the system automatically detects schema updates, validates queries against new schemas, and transforms queries without requiring manual intervention, thereby maintaining API functionality while eliminating time-consuming manual modification processes.
Solution Approach 2:
The patent incorporates feedback mechanisms that automatically detect schema changes in upstream APIs and trigger adaptive responses. The system monitors API schema updates, validates affected queries, and performs transformations based on the detected changes, creating a closed-loop system that maintains functionality while automatically adapting to changes without manual development time investment.
3Manufacturing precision
If separate testing and integration is performed for each API, then integration accuracy is improved, but testing effort and costs increase
Solution Approach 1:
The patent merges separate testing and integration processes into a unified testing framework for the consolidated GraphQL API. Instead of testing multiple separate APIs individually, the system performs integrated testing of the unified API that covers all data sources, maintaining integration accuracy while dramatically improving testing efficiency by eliminating redundant test cases and integration steps.
4Productivity
If distributed tracing is scaled to accommodate growing transaction volume, then system scalability is improved, but complexity of managing tracing processes increases
Solution Approach 1:
The patent segments the distributed tracing management into modular components that handle different aspects of tracing independently. The unified GraphQL API structure naturally segments trace propagation across multiple data sources, allowing each component to manage its own tracing state separately while maintaining overall traceability, thereby scaling transaction throughput without proportionally increasing management complexity.
Data Source
AI summary
Methods and systems for synergistic use of a large language model (“LLM”) and Graph-of-Thoughts (“GoT”) for unified data access across GraphQL application protocol interfaces (“APIs”) via a single API. Methods and systems may include requesting to identify and retrieve data concerning GraphQL APIs from a GraphQL API schema hub. Methods and systems may include identifying key relationships between data elements in the GraphQL APIs using relationship semantic analysis and transporting a text prompt via a cache to an LLM-GOT synergistic processor. Methods and systems may include modeling information generated by the LLM-GoT synergistic processor as an arbitrary graph, generating a unified GraphQL API based on synergistic outcomes, and storing the unified GraphQL API in the GraphQL API schema hub.