UPI is a popular payment instrument , enabling real-time payments that are available round the clock. With growing popularity there is an increased risk of fraud. FSS Sentinel uses machine-learning to monitor UPI transactions in real time and identify anomalous and potentially fraudulent activity.
The system can identify fraud signals to help banks make better informed transaction authorization decisions.
transactions each month
for more than 10 banks
Requiring No Manual
FSS Sentinel’s machine learning technology can internalise subtle changes in payment behaviours and identify emerging fraudulent transaction patterns. This helps banks refine their fraud detection capabilities and stay ahead of new fraud schemes.
Spending patterns with
Customer behavioral profiling
Caters to recent fraud trends , regulatory norms, frequent addition of beneficiaries and frequent changes in PIN
Frequent device changes whether transaction has been initiated from a new device, or one with odd characteristics, such as a foreign keyboard
Demographic and locational changes
End-point device intelligence
Supports Internal, NPCI and third-party scoring models
for Fast Action Against
FSS Sentinel leverages extensive pre-configured rule sets and advanced machine learning to analyse and score every live transaction. The system can analyse transactional and behavioral data from multiple sources to protect against complex fraud attacks. The platform uses over a 100 transaction scoring parameters to score every incoming transaction. Transactions scores are based on a combination of risk elements including:
Frequent UPIN reset
Merchant from where transaction has originated
Multiple UPIN retries
Transactions anomalous to the customer’s transaction pattern
Frequent device and location changes
The system automatically blocks high risk transactions with risk scores exceeding the permissible levels. Transactions from blacklisted or hotlisted VPAs, IP addresses, e-mail and mobile numbers are declined and an alert sent
to the bank for quick action. Real-time risk detection isolates transactions that are more likely to be fraudulent controlling costs associated with fraud loss and reducing the number of follow ups.
Risk Exposure Scores for Improved Security
FSS Sentinel enables merchants to offer their customers multiple payment options to choose from, covering multitude of fraud use cases:
Build Customer Risk Profiles
FSS Sentinel’s machine algorithms can detect online fraud by building customer risk profiles from historic data. The platform can determine if certain activities are typical for a customer profile by comparing unusual spot transaction behaviours with normal behaviours based on:
- Transaction trends – financial/non-financial, type, status
- Customers usage patterns – Host banks PSPs/Other bank PSPs
- Other bank’s customer usage patterns - Host banks PSPs/Other bank PSPs
- Top PSPs adoption trends
Extract Risk Data from All Payment Sources
FSS Sentinel’s adaptors can integrate with structured and unstructured data sources that automatically consolidate information from virtually any enterprise and third-party data source.
Benefits to Banks
Faster Risk Mitigation
Accurate and automated detection which reduces fraud loss, minimizes the time and expense of manual reviews. Machine learning is more effective than humans at detecting subtle patterns to help identify fraudulent transactions.
Gain Customer Loyalty
More customer accounts are protected with pro-active detection of fraud and immediate action.
Time to Market
Banks can go live in a few weeks’ time rather than months as in a Capex intensive on-premise model.
Accommodates Business and Regulatory Change
FSS Sentinel enables banks to quickly and efficiently respond to changes in the global regulatory landscape. Based on compliance risk factors, banks can channel resources at short notice towards high-risk areas.