Dynamic LLM Prompt Adaptation for Open Banking Data Aggregation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing open banking services face challenges in efficiently aggregating and automating data due to the lack of adaptation of artificial intelligence solutions, requiring manual intervention and struggling with the inherent structure and timeliness of open banking data.
Innovation Solution
A dynamic large language model (LLM) is implemented to automatically adapt to open banking system objectives and data-rich vernacular, enabling automated data management and evolution to overcome technological challenges through iterative prompt modifications and retraining.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual intervention is used for data aggregation and decision-making in open banking services, then data processing can be performed with existing technologies, but productivity is reduced and time consumption increases
Solution Approach 1:
The system enables self-service through automated LLM agents that independently perform data aggregation, validation, and decision-making tasks without requiring manual human intervention. The agents autonomously navigate open banking data structures, extract relevant information, and execute service workflows, thereby eliminating time-consuming manual processes while maintaining high productivity.
Solution Approach 2:
Manual mechanical processes are replaced with intelligent automated systems. The patent substitutes human operators with LLM-based agents that use natural language processing and machine learning to perform data aggregation and decision-making, transforming manual mechanical operations into automated intelligent processes that significantly improve productivity and reduce time loss.
2Adaptability or versatility
If existing AI solutions are used for open banking services, then technological interventions can be implemented, but adaptability to open banking data structures and vernacular is insufficient
Solution Approach 1:
The system adapts to open banking data structures by dynamically adjusting parameters such as data validation rules, extraction patterns, and decision thresholds based on the specific characteristics of open banking datasets. The LLM agents are configured with domain-specific parameters that enable them to understand and process open banking vernacular, thereby achieving high adaptability without requiring complex system reconfiguration.
Solution Approach 2:
The LLM-based agent system provides universal adaptability across diverse open banking data structures and service types. Rather than requiring separate specialized systems for different data formats, the patent implements a multi-functional platform where the same core LLM architecture can adapt to various open banking scenarios through configurable prompts and parameters, reducing overall system complexity while maintaining versatility.
3Productivity
If automated technological interventions are implemented in open banking services, then productivity can be improved, but reliability and accuracy of data processing may deteriorate without proper adaptation
Solution Approach 1:
The system incorporates feedback mechanisms where LLM agents continuously evaluate their own performance and adjust their processing approaches. The agents receive feedback on data extraction accuracy, validation results, and decision outcomes, using this information to refine their operations. This self-correcting feedback loop maintains high reliability while preserving the productivity benefits of automation.
Solution Approach 2:
The system performs preliminary actions by pre-configuring LLM agents with domain knowledge, data validation rules, and decision-making frameworks before processing open banking data. This preliminary preparation ensures that automated interventions are executed with high accuracy from the outset, maintaining reliability while achieving rapid service delivery without requiring extensive post-processing corrections.
Data Source
AI summary
A computer-implemented method for providing dynamic LLM open banking services that includes: generating a predefined training action and a predefined prompt modification for merchant entity data prompts and storing with respective metadata configured for matching against values for performance characteristics of the LLM; generating an output based on a first prompt for merchant entity data to the LLM; evaluating the output against the predefined performance characteristics to generate values; matching the values to the predefined training action and prompt modification using the metadata; based on the predefined training action, curating a training data set and retraining the LLM thereon to generate a retrained LLM; based on the predefined prompt modification, generating a second prompt seeking merchant entity data; and generating a second output based on the second prompt to the retrained LLM.