Safeguarding Your Payment System Against Financial Frauds with Fraud Detection and Risk Analysis System (FRS)

By Ellisa | Apr 09 2021 18:30

We are living in a modern world where rapid advancement of technology enables people to conduct multiple financial transactions through a variety of digital channels. As our present-day society opts for more efficiency and convenient portability, digital payment modes such as banking cards, mobile wallets, internet banking, real-time payment, or QRIS, to name a few, have become increasingly preferable to utilize. In fact, the shifting transition from conventional payment methods provides people with a cumulation of varied digital payment solutions that are widely accessible to many, including individuals, corporates and even government institutions.

This advancing evolution is generally followed by the continuous roll-outs of many innovative features, services and benefits that are cohesively designed to facilitate people with an enhanced digital payment experience. It is true that these convenient exchanges of payments create a favourable circumstance that drives resources flow in our economy sector. But factually speaking, they also instigate a crucial challenge, especially in the Indonesia’s dynamic market landscape where regulatory institutions have to tactfully devise the best regulation policies that are not only able to anticipate the rapid growth of payment digitization, but also to ensure that they occur in a safe, secure environment.

With that in mind, we also cannot overlook the fact that acceleration towards digital payments has inevitably attracted many fraudsters out there to attempt numerous detrimental attacks, all targeting any vulnerabilities in the payments system to incur significant losses and exorbitant reputational damage.

“Digital payments landscape is gaining significant traction in Indonesia, yet it consequentially generates lingering fraudulent threats to our digital payments ecosystem. The copious emergence of frauds crimes is often rife with cybercriminals who are constantly scheming for new ways to defraud businesses in a harmful way,” said Patricco Baron, the Chief Technology Officer of PT ALTO Network. “This dire situation reflects the real urgency of how financial institutions should begin to streamline their technology architecture and implement an enhanced analytical fraud detection and risk analysis system. This way, they can respond to fraud with immediate haste, mitigate the risks it causes, and protect any noteworthy assets.”  

Automating Anomaly Detection with Unsupervised Machine Learning and Behavior Anomaly Detection (BAD)

Our financial sector broadly constitutes a wide range of businesses operating at different scales, with different financial service provisions. Starting from local credit unions, commercial banks, to investment companies, for example, it is worth noting that each financial institution has its own unique market and customer segmentation, with a set of behaviour profiles that might be characteristically different to each other. Unfortunately, the burgeoning prospect of Indonesian digital payments ecosystem also induces the rise of differing financial cyberattacks. Fraud threats are constantly coming up with new tactics that are strategically tailored to exploit any particular weaknesses based on one’s unique security system – and when this happens, a financial institution might only be able to protect their own customer circle.

It becomes imperative, then, to develop a robust fraud detection and risk analysis tool with a centralized data system that is capable to quickly identify illicit fraud activities and its multifarious patterns, while simultaneously maintain the value of payment security and data protection to propel the positive growth of our digital economy sector.

However, constructing such system is not easy to execute.

To safeguard against fraud attacks and curb its ever-changing variables, we need a multifaceted, unsupervised machine learning system that is also integrated with a Behavior Anomaly Detection (BAD) feature. Such system will be able to carry out a series of real-time detection and identify any suspicious activities from both internal or external parties, in every transaction layer across the operating network. The system will autonomously anticipate and investigate any anomalies based on learned baseline behaviour profiles, as the network grows and transaction patterns become diverse. It will eventually be able to a derive a unique analysis depending on the acquired information.

Everywhere in the world, security plays a critical element in the digital payments business, including in Indonesia. If we look closely at the Indonesian market, though, identifying unusual deviations with a manual, traditional baselining detection might not be a wise option to take, particularly when we talk about the damaging, costly, and harmful ramifications a single attack can evoke to one’s security system and reputation. Thus, with BAD in implementation, the system will own the ability to accurately distinguish anomalies that do not conform to the normal practices with real-time correlation and reduced error, hence threats can be defused in its early stages.

Screening unauthorized activities faster with ALTO Fraud Detection and Risk Analysis System (FRS)

Addressing numerous financial fraud cases often frequented in the Indonesian digital payments landscape is a challenging task. Building a reliable system to deal with fraud data that keep evolving over time, furthermore, requires more than just tenacity. As a switching and leading finance technology solution provider, ALTO Network resolutely works to create a systemic tool that will essentially strengthen the security layers in the digital finance networks, called ALTO Fraud Detection and Risk Analysis System (FRS).

With an unsupervised machine learning tool, ALTO FRS is designed with the capability to scan every incoming transaction flow and generate a risk score to evaluate the safety of each transaction based on a certain rule or history. The rich dataset collection fed to the system will be centralized and, in turn, benefit the way BAD executes its real-time screening. The engine will attempt to identify, detect, and capture any suspicious or unauthorized activities encountered by the Customers, Terminals, or Banks within the aggregated network. Therefore, the more datasets received by the system, the more sophisticated the scheme of rules that can be applied in its practice.

Not only pointing out the suspicious occurrence, BAD can also help root out problem causes, monitor its progress, and alert the security protection team-in-charge to take action before the attacks manage to further exploit any system vulnerabilities and do real damages.

Innovative features built in ALTO FRS:

  • Real-time detection of any malicious insider or outsider cyberthreats
  • Autonomously distinguish anomalous activities, deviations and outliers from normal practices
  • Unique profile creation for each Customer, Terminal, and Bank
  • Visibility across operating network, system and all application layers for a better activity supervision
  • Systemic adaptability to authorized transaction’s patterns (usage trends, changes or frequencies)
  • Real-time analysis of Customer, Terminal, Banks and other proprietary channels to detect:
    • Smart and customized malwares
    • Social engineering frauds and card scheming attempts
    • Insider misconducts that take advantage of transaction discrepancy


ALTO FRS is fully integrated in every transaction flow and can be deployed to intercept any malicious transactions. It can prevent unwanted security breaches in the system as it continually monitors all types of occurring financial transactions across multiple payment channels such as ATM, Debit, Transfer, Debit Online (eCommerce), Real-Time Payment (BI FAST), Remittance, Biller and QR Payment. Furthermore, the system also enables financial institutions to independently monitor their own respective proprietary channels – which could be the start of a robust and efficient management of fraud risks in the digital era.

“We want to minimize the number of cyber fraud cases that keep hindering the growth of Indonesia’s digital payments ecosystem. And ALTO FRS is created with that purpose,” Patricco Baron said further. “ALTO FRS is capable to scan and comprehensively analyze all incoming transaction profiles and its contained variables. It will learn any inherent derivations resulting from the growth of transaction volumes or the evolution of usage patterns, and form a behavioural baseline that can be utilized immediately to perform the necessary countermeasures.”

In upholding a strong relationship that inspires trust and assurance, Patricco Baron believes that a financial institution has a chief responsibility to its customers and businesses to protect itself against the fraudsters. To help achieve this objective, ALTO Network will provide only the best security protection system that enables our financial institution members to safeguard their systems against potential risks and reputational damage induced by the frauds; while simultaneously putting a fundamental importance on data privacy.

With an extensive experience in the field of financial security practice, ALTO Network truly understands the value of data privacy markedly held by each of our members. While part of ALTO FRS’ operation includes centralizing all incoming datasets in order to perform an exceptionally smarter anomalous activity detection, the aforementioned datasets – to be precise – are strictly and exclusively limited to customer behaviour data and fraud profiles. As such, ALTO never shares the specific transaction data, such as our members’ transactional data or its customers’ personal data, to anyone including our other members under any circumstances. This principle is incorporated into each of ALTO’s operational procedure, to effectually build a positive credence and trust within the digital payments community.


More ALTO Daily articles