Burak Yildirim, Team Lead, Adyen, writes in a blog post that being the most “money-involved domain” of Fintech, payment processing platforms such as Adyen are frequently targeted by “individuals or sophisticated organizations.” Fraudsters or bad actors often try to generate “synthetic” identities such as hundreds of different email addresses or even spoofed IP addresses “hoping to slip-in” to payment platforms or systems.
However, Yildirim noted that with the assistance of their in-house Adyen Graph Database, they are able to effectively track these malicious attempts in “real-time” and can also block these fraudulent activities “even though they keep trying new tricks.”
Yildirim explains that Adyen Graph Database is one of the company’s in-house products which assists them with identifying and catching fraudulent actions or attempts. It also helps with identifying individuals or businesses attempting to open an account in Adyen. Yildirim points out that you don’t want to carry out business activities with “shady companies or sole proprietors.”
Going on to explain how the graph data structure may be helpful with payment data, Yildirim notes that use cases may include catching fraudulent activities or identifying a business or an individual during their onboarding to Adyen and “continuously checking if they comply with Adyen and payment regulations, pattern matching is a must.”
“To see and explain how payments are connected to each other, relational data structures are not the best choice on their own. There can be changing patterns, many hops between fake identities, recursive and exhaustive search helps us to see all the patterns and connections in real-time, enabling us to track criminals such as fraudsters or shady business owners attempting to open an account and start to process payments with Adyen.”
While discussing why Adyen has its own graph database solution, Yildirim noted that with the payments data, you need to be able to look at that from different yet “combined angles: as a transactional time-series data and as a connected graph to closely track potential criminals.”
He further noted that having a pure traditional relational database system may not help as they are usually not good enough at “handling connections and patterns.” However, graph databases aren’t well-suited for processing “high volume transactional data.”
“At Adyen, we need a completely different solution that combines both properties of graph databases and relational-transactional databases in a way that they also serve our specific requirements such as: transactional (because of the properties of payments data).”
Yildirim also mentioned that Adyen handles thousands of different firms or businesses’ payments with different business requirements. Also, for compliance reasons, it “does make sense to process each company’s data differently.”
While sharing more details on their approach, Yildirim pointed out the importance of maintaining data privacy: “as a global financial institution, we encrypt any PII (personally identifiable information) data before processing and persisting.”
The Adyen team concluded:
“With the exponential growth in our processed payments volume, we constantly look for improving our bespoke graph database solution to always scale better, be more extensible without scarifying the ability to be always ahead of the game before criminals as they constantly try new ways to penetrate into payments domain.”
(Note: to learn more about engineering at Adyen, check here.)
As covered in December 2020, Adyen revealed how companies are simplifying and streamlining digital payments with its products.