According to a company representative, as digital payments rapidly expand, financial fraud schemes have grown more sophisticated. This trend demands that financial institutions deploy technology capable of real-time risk detection while ensuring a stable transaction experience for customers.

Addressing this need, BPC developed the SmartVista fraud management solution. This platform utilizes artificial intelligence (AI) and machine learning (ML) to handle increasingly diverse fraud scenarios. The system employs a comprehensive approach, enabling real-time decisions across multiple transaction channels, including cards, accounts, and digital payment methods. This allows financial institutions to monitor and detect anomalies as they occur.

Customers using the SmartVista solution. Photo: BPC

Beyond real-time processing, the system also supports near real-time and offline analysis. These tools are crucial for investigations and risk management. The platform combines AI with link analysis to identify anomalies across various transaction types, including point-of-sale card transactions, online transactions, mobile banking, internet banking, and e-commerce.

The platform's omnichannel capability allows monitoring across diverse payment areas: from acquirers and issuers, token provisioning, QR payments, and e-commerce to mobile banking, internet banking, e-wallets, account-based payments, and instant payments. This enables financial institutions to build a unified 360-degree view of all financial and non-financial events throughout their system.

According to a BPC representative, instead of relying on fixed rules or manual checks, SmartVista integrates multiple methods within a single processing framework. The system simultaneously employs rules, list checks, behavioral analysis, and machine learning-based risk scoring. This approach allows each transaction to be evaluated based on factors like user identity, transaction channel, and consistency with normal behavior. Consequently, the system can detect suspicious transactions early and prevent risks before they materialize, while ensuring valid transactions are processed quickly.

The platform uses both supervised and unsupervised learning models to identify fraud. These models help the system adapt to changes in transaction behavior over time. Additionally, the system can suggest new rules based on emerging fraud patterns, shortening reaction times to new attack methods. Operational teams can also adjust rules and processes directly within the system, enhancing control efficiency.

SmartVista's low-code architecture enables organizations to quickly test and deploy fraud prevention scenarios. The system supports adding authentication steps, such as biometrics, when necessary. This flexibility is crucial given the continuous evolution of fraud methods and increasingly updated compliance requirements.

Customers paying with QR codes. Photo: BPC

A notable feature of the platform is its graph-based link analysis capability. This feature helps financial institutions move beyond individual alerts to identify signs of organized fraud. The system uses a graph database to build and analyze relationships among various entities like customers, cards, accounts, devices, and payment acquirers. From this, analysts can visually observe linked clusters, central points, and unusual connections directly on the interface.

This approach aids in detecting intermediary account networks, organized fraud schemes, and hidden connections between transactions. Furthermore, an integrated AI assistant supports investigations by providing additional context for each alert and suggesting appropriate handling. As a result, financial institutions can enhance fraud detection efficiency while building consistent and scalable control processes during operations.

SmartVista can be deployed flexibly on-premise or in the cloud. This allows financial institutions to comply with local data storage regulations while leveraging system upgrades, data analysis capabilities, and performance scalability.

BPC reports deploying the SmartVista platform for numerous financial institutions across the Asia-Pacific region, Europe, and other markets. In Malaysia, Co-opbank Pertama (CBP) transitioned from manual verification processes to a near real-time, cloud-based fraud monitoring system. The bank also integrated machine learning models, utilizing historical fraud data to improve detection capabilities. BPC provided training to help operational teams effectively use the system's features. Consequently, CBP met Bank Negara Malaysia's regulatory requirements and strengthened trust in its digital services.

In Pakistan, Meezan Bank and Samba Bank implemented SmartVista to enhance fraud control across their digital ecosystems. At Meezan Bank, the system established a unified AI-based control framework, encompassing multi-channel monitoring, real-time detection, and regulatory compliance support. The platform combines AI/ML analysis, link analysis, and flexible rule-setting tools. It also sends alerts via multiple channels, such as SMS, email, and app notifications, increasing transaction transparency.

Following implementation, the bank prevented approximately 1,5 million fraudulent point-of-sale transactions with significant total value, while also detecting and blocking numerous e-commerce attacks. Throughout this process, the false positive rate remained low, minimizing impact on customer experience. The results demonstrate the system's ability to scale with the growth of digital channels while maintaining operational efficiency.

In Europe, Bulgaria's DSK Bank consolidated its fraud control systems into a single platform, significantly reducing operational costs across transaction channels. In many markets, financial institutions are observing similar trends as they shift to AI-based fraud detection models. These systems help mitigate risks and enhance trust in digital financial services.

Overall, SmartVista offers comprehensive monitoring, real-time analysis, and flexible configuration. This empowers financial institutions to strengthen controls against increasingly sophisticated fraud, maintain stable operations, and improve operational efficiency.

Hoang Dan