AI Payment Recommendation Engine for Fraud-Approval Trade-offs

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Issuers and merchants face challenges in real-time decision-making to maximize approval rates and minimize fraud for card-not-present (CNP) payment transactions, as they struggle to determine which authorization decision products to apply.

Innovation Solution

A server system equipped with a data-processing engine, a reinforcement learning (RL) agent, and a product recommendation engine processes payment authorization requests in real-time, identifying optimal combinations of authorizing components using a trained machine learning model based on transaction features.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If multiple authorization decision products are applied to CNP transactions to improve security and reduce fraud, then fraud detection capability is improved, but approval rate deteriorates due to increased transaction declines

Engineering Contradiction:

Improvefraud detection capabilityVSAvoidapproval rate

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The system dynamically changes the parameters of authorization decision-making by using machine learning models to select and combine different authorization products based on transaction characteristics, cardholder behavior patterns, and risk factors, optimizing the balance between fraud detection and approval rates

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The system implements dynamic authorization strategies where the combination of authorization products is not fixed but adapts in real-time based on the specific transaction context, cardholder history, and detected risk patterns, allowing flexible adjustment between security and approval

Inventive Principle:

Principle #15Dynamics

2Reliability

If fraud scoring models are used to secure CNP transactions, then security is improved, but legitimate transactions deteriorate due to false positives and declines

Engineering Contradiction:

ImprovesecurityVSAvoidlegitimate transaction approval

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The system incorporates feedback loops where authorization outcomes, fraud detection results, and transaction patterns are continuously fed back into the machine learning models to refine scoring accuracy and reduce false positives, improving both security and legitimate transaction approval

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The system performs preliminary risk assessment and pre-selection of appropriate authorization products before the final authorization decision, allowing legitimate transactions to be identified early in the process and routed through optimized authorization paths that minimize false declines

Inventive Principle:

Principle #10Preliminary action

3Measurement precision

If issuers manually decide which authorization products to apply to each transaction, then decision accuracy is improved, but processing time deteriorates due to real-time decision complexity

Engineering Contradiction:

Improvedecision accuracyVSAvoidprocessing time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system replaces manual mechanical decision-making with automated machine learning models that analyze transaction features and select authorization products algorithmically, achieving high decision accuracy without human intervention and eliminating processing delays

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system enables self-service authorization decision-making where the machine learning models autonomously select and apply appropriate authorization products based on transaction characteristics without requiring manual issuer input, maintaining accuracy while reducing processing time

Inventive Principle:

Principle #25Self-service

Data Source

PatentUS12293362B2Artificial intelligence based product recommendation methods and systems for enhancing approvals of payment processing requests

Publication Date: 2025.05.06 MASTERCARD INT INC

AI summary

Recommendations of one or more authorizing components are provided to issuers and/or merchants for enhancing approval rates of payment processing requests. A server system receives a payment authorization request for a payment transaction between a cardholder and a merchant in real time. Payment transaction features associated with the payment transaction are identified based on the payment authorization request. A combination of one or more authorizing components to be applied to the payment transaction is predicted to obtain a product recommendation strategy for the payment transaction. The combination of one or more authorizing components is predicted based on a trained machine learning model and the payment transaction features. The payment authorization request and the product recommendation strategy are transmitted to an issuer associated with the cardholder.