Smart Rules

Smart Rules allow merchants to create multi-condition fraud prevention systems across industries such as e-commerce, digital goods, travel, jewellery, utilities, and more. The rule engine evaluates parameters, including transaction details, device signals, geographic data, payment modes, and other factors, to block fraudulent activity while allowing legitimate transactions to process seamlessly.

​

Why choose smart rules

Smart Rules offer the following benefits:

• Comprehensive coverage: Over 30 risk parameters are evaluated in real time across devices, locations, identities, payments, and user behaviours to support advanced fraud detection strategies.
• Flexible logic engine: Supports advanced operators such as range-based filtering, multi-value matching, and nested AND/OR conditions, enabling precise and customisable rule creation.
• Business-friendly interface: Provides a visual rule builder with plain-language explanations and a pre-built rule library, enabling quick deployment without technical expertise.
• Adaptive risk management: Rules can be updated and refined in real time based on emerging fraud patterns, ensuring continuous protection without disrupting legitimate transactions.
• Scalability: Handles high transaction volumes without performance degradation, making it suitable for fast-growing or high-risk businesses.

​

Use cases

You can apply Smart Rules to the following scenarios for threat protection:

​

Smart rule logic and configuration parameters

Smart Rules are built using a simple When–Then logic structure, where conditions are defined under When and corresponding actions are executed under Then. There are two types of Smart Rules:

​

Conditional rules

These rules are executed in real time, based on specific conditions and fixed parameters. Example 1: If the transaction amount is ≥ ₹50,000 AND the proxy type is VPN AND the city is Jamtara, then block further transactions. Example 2: If the transaction amount equals ₹10,000 OR the UPI VPA provider is Mobikwik, AND the city is Jamtara, then the system blocks further transactions to prevent potential fraud.

​

Conditional rule explanation

The above rule is designed to block transactions when specific high-risk conditions are met. WHEN

• Amount = 10,000

OR

• UPI VPA Provider = Mobikwik

AND

• City = Jamtara

THEN

• Block further transactions.
• The user will be unable to complete this transaction or any other transaction that meets these conditions.
• No debit occurs from the user’s account.

The system checks if either the transaction amount is exactly ₹10,000 or the UPI VPA provider is Mobikwik. If this is true, and at the same time the transaction originates from the city of Jamtara, then the transaction is blocked, meaning the user cannot complete it.

​

Aggregate rules

These are aggregated rules that consider a user’s past transactions before triggering actions. They combine velocity checks with multiple conditions to create optimised, comprehensive rules. Example 1: Limit the total successful transaction amount to ₹25,000 per card per hour. Example 2: If a single card number is linked to more than three mobile numbers in the past 24 hours, the system blocks further transactions to prevent potential fraud.

​

Aggregate rule explanation

The above rule is designed to prevent card sharing among users by preventing more than three mobile numbers from being linked to the same card for payment in a given day. WHEN

• For each Card Number, the Number of linked parameters i.e Mobile Number is more than 3 within the last Day.

THEN

• Block further transactions.
• Prevents linking a fourth or additional mobile number to the same card for payments in that time period.

​

Rule outcomes

The two Smart Rule outcomes are:

• Block further transactions: Prevents any new transactions from being processed. Users will not be able to complete a transaction, and no debit will occur. In UPI Intent cases, a debit may still happen but is immediately refunded. All such attempts are marked as FAILED in Cashfree’s system.
• Flag further transactions: The transaction is processed like any regular transaction, and the merchant receives settlement as usual. However, the transaction is highlighted in the Risky Transactions section of RiskShield for monitoring and review.

​

Smart rule parameters

Smart Rules evaluate transactions using the following parameter types to identify and prevent potential fraud attempts:

​

Conditional rule parameters

The following table lists the parameters available for creating conditional rules:

​

Aggregate rule parameters

The following tables lists the parameters available for creating aggregate rules:

​

Aggregation setup parameters

​

Filter criteria parameters

​

Create a smart rule

Follow these steps to build a Smart Rule based on transaction criteria:

1. Log in to the Merchant Dashboard.
2. Navigate to RiskShield > Smart Rules.
3. Click Create Smart Rule.
4. Enter a descriptive Rule Name.
5. Configure the When conditions:
   1. Select the type of filtering from the dropdown menu, such as Condition or Aggregate based.
   2. Select the parameter used to evaluate transactions on.
   3. Select the specific operator based on the chosen parameter.
   4. Enter the comparison value(s).
   5. Click Add Condition Set for additional conditions with AND/OR logic.
6. Configure the Then outcome: Block or Flag further transactions.
7. Review the auto-generated rule Explanation in English for verification.
8. Click Confirm Rule to create the new rule.

​

Managing existing smart rules

Once created, all configured Smart Rules appear in a comprehensive dashboard table under the Smart Rules section, providing complete visibility and control over your fraud prevention system.

​

Rule dashboard overview

The Smart Rules dashboard displays three main categories:

• All Rules: Complete list of all configured rules which includes activated and deactivated rules.
• Active Rules: Currently enabled and executing rules.
• Recommended: Pre-configured rules suggested by Cashfree for fraud prevention use.

The rule status indicators available are:

• Active: The rule is currently enabled and actively monitoring transactions.
• Inactive: The rule is disabled and not monitoring transactions.
• Recommended: The rule is suggested by Cashfree based on transaction patterns and risk factors.

​

Scenario examples

The following examples show practical Smart Rule configurations with real-world scenarios. Each example illustrates the conditions, actions, and outcomes to guide effective fraud-prevention strategy setup.

​

Conditional rule-based scenarios

The following scenarios illustrate common situations where conditional rule-based filtering enhances fraud detection and transaction control.

​

Aggregated rule-based scenarios

The following scenarios highlight common use cases where aggregate rule-based monitoring helps detect abnormal patterns and prevent fraudulent activity.