Feedzai fights fraud using machine learning and big data analytics. The San Mateo startup, founded in 2011, partners with payment processors, banks and merchants to identify criminal activity across s
Feedzai fights fraud using machine learning and big data analytics. The San Mateo startup, founded in 2011, partners with payment processors, banks and merchants to identify criminal activity across sales channels. Major clients include First Data, Celfocus, Coca-Cola, Logica, Vodafone, Ericsson, and Servebase Credit Card Solutions.
As blockchain technology gains attention, Feedzai has begun exploring how it could strengthen fraud detection systems. Paulo Marques, the company's founding partner and chief technology officer, discussed the potential in a recent interview.
Blockchain creates permanent transaction records that operate without a central authority. This setup offers meaningful advantages. Participants in the network can verify transactions themselves rather than trusting a single institution to maintain accurate records. "No one entity controls blockchain," Marques says. "Every participant can verify the transactions, and you are not forced to rely on one entity (like a bank) to keep track of balances." The system also cuts costs for both organizations and consumers. By removing dependence on intermediaries and building in transparency, blockchain makes tampering difficult.
The permanence and decentralization protect transaction histories. Blockchain can prevent certain attacks, such as double spending. It cannot, however, detect fraud or predict criminal behavior without additional tools. Marques compares the situation to online payments. Credit cards made e-commerce possible, but the technology also introduced new risks. "Using a credit card to make an online purchase is possible because of digital currency," he notes, but the shift opened doors to different fraudulent schemes.
Blockchain has inherent limitations. It provides no defense against account takeovers, identity theft, or financial scams. In fact, criminals can exploit blockchain for money laundering and other illicit purposes. These gaps require additional safeguards.
Machine learning fills that role. Feedzai layers its algorithms on top of blockchain to enhance fraud protection. The software identifies behavioral patterns unique to each user, which the company describes as "commerce DNA." Marques notes that "a fraudster can steal your data, but they can't steal your behavior."
Fraudsters evolve their tactics faster than traditional systems can react. Marques points out that rule-coded approaches cannot keep pace. "Today's fraud patterns are evolving much too quickly for human-coded, rules-only systems to operate effectively," he says. Feedzai's algorithms watch transactions as they occur and adjust their understanding in real time. When applied to blockchain infrastructure, this approach creates a solution more comprehensive than blockchain alone can provide. The combination may even predict criminal behavior before it happens.
Feedzai does not develop blockchain technologies. Instead, the company supplies machine learning algorithms that enhance existing blockchain systems by adding a fraud-detection layer that neither technology provides on its own.
