Open Banking LLM Adaptation for Prompt and Retraining Feedback

Resolve Bottlenecks,
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, leading to manual intervention and inefficiencies in managing diverse data structures and timeliness requirements.

Innovation Solution

A dynamic large language model (LLM) is implemented to automatically adapt to open banking system objectives, enabling efficient data aggregation and analysis by iteratively refining prompts and training data to improve performance characteristics.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If manual intervention is used to aggregate and manage open banking data, then data accuracy and control are improved, but productivity and timeliness deteriorate

Engineering Contradiction:

Improvedata accuracyVSAvoiddata aggregation efficiency

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The LLM system performs self-learning and self-optimization by automatically evaluating its own outputs against performance characteristics and iteratively improving through self-generated training actions and prompt modifications, eliminating the need for manual intervention while maintaining high data accuracy

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent replaces manual mechanical processes with an automated LLM-based system that uses natural language processing to aggregate, validate, and manage open banking data, significantly improving productivity while maintaining reliability through intelligent automation

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Extent of automation

If existing AI solutions are used for open banking, then automation is improved, but adaptability to diverse data structures and vernacular deteriorates

Engineering Contradiction:

Improveautomation levelVSAvoiddata structure adaptability

Core Design Contradiction:

Extent of automationVSAdaptability or versatility

Solution Approach 1:

The LLM system is dynamically adaptable to different data structures and vernacular through continuous learning and iteration, automatically adjusting its behavior based on the specific characteristics of incoming open banking data from diverse sources

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system changes its processing parameters and approaches based on the specific data structure and vernacular encountered, allowing it to adapt to diverse financial data formats and language styles from different institutions and jurisdictions

Inventive Principle:

Principle #35Parameter changes

3Loss of time

If LLM is statically trained, then development time is reduced, but performance optimization capability deteriorates

Engineering Contradiction:

Improvedevelopment timeVSAvoidperformance optimization

Core Design Contradiction:

Loss of timeVSReliability

Solution Approach 1:

The system implements continuous feedback loops where LLM outputs are evaluated against predefined performance characteristics, and this feedback drives automatic retraining and optimization, ensuring continuous performance improvement without extending development time

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The LLM system operates continuously with ongoing optimization, maintaining useful action throughout operation rather than stopping for periodic retraining, ensuring both rapid deployment and continuous performance improvement

Inventive Principle:

Principle #20Continuity of useful action

4Reliability

If iterative LLM retraining is performed, then performance characteristics are improved, but computational resource consumption increases

Engineering Contradiction:

ImproveLLM performanceVSAvoidcomputational resource consumption

Core Design Contradiction:

ReliabilityVSUse of energy by moving object

Solution Approach 1:

The system performs partial retraining using only the specific training actions and prompt modifications identified as needed through evaluation, rather than complete retraining, reducing computational resource consumption while maintaining performance improvements

Inventive Principle:

Principle #16Partial or excessive action

Solution Approach 2:

The iterative optimization focuses computational resources on specific areas where performance improvements are needed, identified through evaluation against predefined characteristics, rather than uniformly processing all data, thereby reducing overall resource consumption

Inventive Principle:

Principle #3Local quality

Data Source

PatentUS20260050972A1Computer-implemented methods, systems comprising computer-readable media, and electronic devices for providing financial network large language model dynamic open banking services

Publication Date: 2026.02.19 MASTERCARD INT INC

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 value optimized transaction 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 value optimized transaction 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 value optimized transaction data; and generating a second output based on the second prompt to the retrained LLM.