Payment Authorization Routing with Machine-Learned Request Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing payment routing systems are static in their logic, leading to a high rate of declined payment authorization requests, resulting in lost revenue, customer dissatisfaction, and unnecessary network communications, with negative downstream consequences such as fraud screening and increased computing resource consumption.
Innovation Solution
A computing system utilizing machine learning to predict success probabilities for payment authorization requests by generating multiple candidate requests with varying parameter combinations, selecting the most likely to succeed, and optimizing routing to improve authorization rates and reduce resource consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a static payment routing system is used, then the system structure is simple and easy to implement, but the payment authorization success rate is low and many requests are declined
Solution Approach 1:
The patent applies dynamics by transitioning from a static routing system to a dynamic machine learning-based routing system. The ML model continuously learns from historical payment data and adapts routing decisions in real-time based on patterns in merchant categories, geographic locations, payment methods, and timing, thereby improving authorization success rates while managing complexity through automated learning.
Solution Approach 2:
The patent replaces the mechanical/static routing logic with a machine learning system that uses computational models to predict optimal routing decisions. The ML model substitutes traditional rule-based routing mechanisms with data-driven predictions, enabling more accurate and adaptive payment authorization routing.
2Reliability
If multiple retry attempts are made after a declined authorization, then the chance of successful payment increases, but computing resources and network communications are wasted
Solution Approach 1:
The patent applies preliminary action by using the machine learning model to predict the likelihood of successful authorization before attempting payment routing. This allows the system to pre-determine the best routing option and avoid unnecessary retry attempts that would consume computing resources and network communications, thereby reducing energy loss while maintaining payment completion rates.
3Adaptability or versatility
If the same static authorization request parameters are used consistently, then the routing logic is simple, but the adaptability to different payment scenarios is poor
Solution Approach 1:
The patent applies parameter changes by using the machine learning model to dynamically adjust authorization request parameters such as routing selection, merchant category codes, geographic routing, and payment method preferences based on the specific characteristics of each payment scenario. This enables the system to adapt to diverse payment situations while the ML model manages the complexity of parameter optimization.
Data Source
AI summary
Computing systems and methods can use machine learning to improve the authorization of payments by payment systems. Specifically, contrary to existing payment authorization approaches which are static in nature, example aspects of the present disclosure are directed to machine learning systems which enable the dynamic and real-time optimization of one or more variable request parameters associated with a payment authorization request. Specifically, example computing systems described herein can employ one or more machine-learned models to assist in selection of a particular payment processor to which the authorization request is routed, optimization of one or more variable message parameters included in the authorization message (e.g., selection of values for a merchant identification, a merchandise category code, a transaction type, and/or other variable message parameters), and/or automatic generation and/or execution of an automated retry strategy that can be executed if a first authorization request is declined.