Each year, online merchants and banks lose billions of dollars to fraud. The cost of managing US$1 of fraud loss is US$3.52, with fraud costs accounting for up to 6 per cent of a merchant’s total annual expenditure, according to SG:D accredited company CashShield, a Singapore company that now uses machine learning to detect fraud and prevent it.
The system is fully automated and relies on machine learning to eliminate the need for manual reviews. So confident is the company of its algorithm that it offers to a 100 per cent chargeback to customers who are scammed by fraudsters.
While it is not the only one with such automated fraud detection technology, the difference is its “Wall Street-style, high-frequency trading approach” that determines how risky a transaction is. This enables it to determine if the transaction should go through.
A real-time pattern recognition technology also enables CashShield to detect possible fraud cases. This works even for new transactions that do not have historical data for a machine to be trained on. It is among the key differences that CashShield has impressed customers such as Alibaba and Razer with.
It has helped many businesses simplify their operations for greater efficiency while making fraud detection a much less expensive endeavor. Merchants care about revenues, reducing friction in the customer user experience, streamlining operations, scalability of processes and resource efficiency, and CashShield has delivered these benefits.
Besides e-commerce, the technology can also be applied to other areas. For example, social media accounts that are often compromised can be better safeguarded if the system can identify a suspicious login and prevents it from going through as usual.