Guilherme Chaddad, Software Engineer at Pismo, was a panellist at a Sumsub webinar about transaction monitoring. He joined Sumsub’s CTO, Vyacheslav Zholudev, to discuss how this process helps companies detect suspicious activity, perform risk assessments, and minimise financial crimes such as money laundering and fraud. Lucas Newman, Business Development Manager at Sumsub, moderated the conversation.
Transaction monitoring is crucial for companies providing financial services. At Pismo, we partner with Sumsub for identity orchestration, fraud prevention, and KYC/KYB & AML compliance. This enables banks and fintech companies to monitor events when users sign up for a service, attempt to pass KYC, log in to a service, change passwords or 2FA codes, or make a payment, for example.
Monitoring behaviours
Pismo processes financial transactions like deposits, withdrawals, money transfers, and purchases with credit and debit cards. Guilherme explains that monitoring those activities is vital at account and platform levels.
“At the account level, we look into metrics like the transaction volume, the location of the establishment where a purchase occurs, and the account holder’s behaviour. When something doesn’t fit the customer’s transaction history, our monitoring system detects the odd behaviour and issues an alert,” he says.
The same occurs when Pismo monitors transactions at the platform level. According to Guilherme, transaction approvals and denials usually follow a pattern. Still, it’s common to see temporary increases in the transaction volume – which does not mean these transactions are not legitimate.
“It’s normal to have sudden rises in the load when there is a promotional sales campaign, for example. If the transactions are legitimate, approvals and denials keep the regular pattern. We suspect there may be a fraud attempt when, for instance, there is a swelling in operations with accounts that are not in the platform’s database.”
Automated versus manual reviews
When setting up a transaction monitoring process, Pismo’s team establishes rules to help detect fraud and suspicious activity. The more rules we have in place, the more fraud and suspicious activity we can detect. However, creating more rules also increases the number of false alarms. It’s a delicate balance for the customer experience—keeping their account safe or potentially creating false alarms.
For effective transaction monitoring, balancing automated and manual reviews is crucial. Guilherme says automatic checks allow Pismo to create rules, trigger alerts, and block suspicious accounts, but these actions require manual reviews.
“Our team reviews the alerts so that we can evaluate if they are false positives or if we have to tweak our rules or improve the user experience somehow. Good manual reviews and analyses of the dashboard and collected data help improve the automated reviews,” he says.
KYC and transaction monitoring
It’s possible to combine Know Your Customer (KYC) tools with transaction monitoring to improve it. Pismo uses data from customer onboarding, such as age, country of residence, nationality, answers to questionnaires, and risk labels (risky email, virtual phone number, usage of VPN, etc.) to automate routines like requesting a confirmation on high-volume transactions, checking the funds’ source or even blocking an account.
“Everything starts when a customer is onboarded. KYC tools help us set the profile where we define limits and spending rules for a specific account. Based on this information, we build monitoring tools, automated validations, and spending control rules. Then we set the strategies to enable the operations,” Guilherme concludes.
To learn more about the importance of transaction monitoring and its role in fraud prevention, be sure to watch the full webinar, available on-demand.
