ARDID performs real-time checks related to the reputation of the transaction that a client attempts to make.

Both positive and negative reputations are easily configured and assigned a weight. Each reputation is individually evaluated and then totaled, based on which the transaction's risk level is determined. Similar to Machine Learning rules, this can result in:

• Acceptable risk (approve the transaction)

• Low risk (request 2FA authentication)

• Moderate risk (reject the transaction)

• High risk (block the client's account)

Among the variety of available reputational rules are:

• Known geolocation
• Suspect geolocation
• Known IP
• Suspect IP
• Known account identification
• Suspect account identification
• Known recipient
• Suspect recipient
• Known email domain
• Suspect email domain
• Known email
• Suspect email

The rules are defined and activated or deactivated on the fly, without the need to write code or restart services.

As many atomic rules as necessary can be defined, and if applicable, all will be analyzed, and the most restrictive decision will be made.