Payment Risk Platform Core Features

This page provides an overview of the core features of the Payment Risk Platform (PRP). The following features are available for both managed and self-service PRP.

Feature Platform

The Feature Platform is a platform for you to easily use real-time data transformers and aggregators as part of the Galileo Risk Platform, allowing you to more efficiently manage your data for fraud prevention. Feature Platform enables you to spend less time on the risk feature generation and to focus more on implementing your fraud detection and prevention strategies in the Decision Rules Engine.

Feature Platform supports all of the following:

• Use of out-of-the-box features that leverage Galileo’s domain expertise for different risk and fraud use cases.
• Ability to build, test, and deploy features using the UI or by programming code snippets on PRP.
• Create feature templates which can be assigned to feature packages to be shared with other PRP users.
• Ability to generate features in batch on a regular basis, or in real-time, low latency, high Transaction Per Second (TPS) production environment.
• Adjust features based on historical production data

Decision Rules Engine

The Decision Rules Engine is complementary to the Feature Platform. The Decision Rules Engine empowers you to create fraud rules using the available features in the Feature Platform to make fraud decisions in real-time or in a batch mode. When a suspicious event or user is detected by a rule in Rules Engine, a response in a pre-configured approved format is sent to you which includes the rule name, a rule note, and a fraud decision recommendation.

Like the Feature Platform, you can build, test, and deploy rules directly in PRP. Rules can be evaluated in a certain sequence using Decision Flows and configured at different levels of the customer journey for optimal fraud decisions. You can have one or more decision flows, with each decision flow mapped to one or more use cases.

Use case

A program manager notices low-dollar, high-velocity transactions from various Google merchants. Based on your fraud team investigation, you decide to deploy a real-time fraud rule to decline the same patterns going forward. Using PRP’s Feature Platform, the fraud team configures a new velocity feature to target this specific fraud trend. The velocity feature monitors historical transactions on the specific account for the last 30 minutes in a rolling window and counts the number of transactions at Google merchants. Your fraud team also configures a rule in Rules Engine based on the combination of the velocity features and the Transaction Risk GScore. Rules Engine sends the fraud decision recommendation in real-time when the next transaction with the same condition hits PRP.

How it works

Here is a brief overview of how Feature Platform and Rules Engine are used to solve the use case outlined in the section above:

1. A fraud agent configures a rule to block the low-dollar, high-velocity transactions at Google merchants.
   • Rule Name: "hr_google_trnx”
   • Rule Action: “DENY”
   • Rule Condition: IF the merchant/acquirer description contains “Google” AND Galileo Transaction Risk GScore >= 0.9 AND the number of transactions at Google merchant from the same account for the last 30 minutes >= 5 THEN trigger the Rule Action.
2. PRP identifies a new transaction with similar conditions.
3. PRP evaluates the transaction against the rules defined in the Rule Engine.
4. Rule Engine returns the Rule Action “DENY” along with the Rule Name “hr_google_trnx” to you.
   • If you are using self-service PRP, this data is returned via the Risk API.
   • For managed PRP clients, this data can be returned via denied_auth event if you use the Events API, or the Auth API.

Case Management System

The Case Management System is a post-fact fraud trend monitoring tool that uses all transaction data on PRP, such as transactions, accounts, and review history. The Case Management System uses machine learning to enable a prioritized review queue based on risk score task assignment. The Case Management System is tightly integrated with Knowledge Graph, Galileo’s visualization tool for the linkage analysis and for the fraud ring investigation. Knowledge Graph can be used for individual or groups of transaction data. It connects transaction data and analyzes the relationship to uncover patterns in near real-time.

How it works

To review suspicious transactions that are captured by the fraud rules in Rules Engine:

1. Your fraud team creates the review queue logic and defines Case Management queue logic in the Rules Engine.
2. PRP identifies a new transaction that satisfies the review queue logic.
3. Transaction information is automatically sent to the designated queue in the Case Management System and assigned to a designated fraud agent.
4. The designated fraud agent is alerted of a new review task assignment.
5. The agent checks the transaction and the account details based on your Fraud Ops procedure.
6. Agent marks the transaction, the account, and the individual as confirmed fraud and assigns them to a black list to prevent further fraudulent transactions.

Request access to the Payment Risk Platform Self-Service User Guide for details on how to implement the features outlined in this guide.