Featurespace Blocks Fraud Attacks for Financial Institutions with AI and NVIDIA GPUs

E-commerce sales have soared as more people shop remotely, spurred by the pandemic. But the push has also led fraudsters to use the opportunity to rip off retailers and customers, according to David Sutton, chief analytics officer at fintech firm Featurespace.

The UK-headquartered company has developed AI-based technology to increase the speed and accuracy of fraud detection and prevention. Called ARIC Risk Hub, the platform uses deep learning models trained using NVIDIA GPUs to distinguish between valid and fraudulent transactional behavior.

“Online transactions are a prime target for criminals because they don’t need to have the physical card to transact,” Sutton said. “With compromised card details readily available via the dark web, fraudsters can target large volumes of cards to commit fraud with very little effort.”

ARIC Risk Hub creates complex behavioral profiles of what it calls “authentic” customers by converging third-party transactions and data throughout their lifecycle within a financial institution.

Fraud prevention has traditionally been limited by detection delays, with customers not being notified until after the money has already left their bank accounts. But ARIC Risk Hub in less than 30 milliseconds identifies anomalies in even the smallest changes in a client’s behavior. It compares each customer’s financial event to their profile using adaptive behavioral analytics powered by AI.

The technology is deployed at 70 major financial institutions around the world – and some have reported that it has blocked 75% of its fraudulent attacks, Sutton said.

ARIC Risk Hub helps these institutions identify criminal behavior in near real time, reducing their financial losses and operational costs, and protecting more than 500 million consumers from fraud and financial crime.

Featurespace is a member of Creation of NVIDIAa free global program that encourages cutting-edge startups.

100x Model Training Acceleration with NVIDIA GPUs

Featurespace got its start over a decade ago as a machine learning consultancy. It was rooted in the research of Professor Bill Fitzgerald of the University of Cambridge, who sought to make a business impact with adaptive behavior analysis, a technology he created.

Applied to the financial services industry, the technology has quickly taken off.

“With this technology, you can build a deep learning model that learns and understands the kinds of actions a person normally takes so that you can look for changes in those actions,” Sutton said.

Previously, it took Featurespace weeks to configure and train different deep learning models. With NVIDIA A100 Tensor Core GPUthe company has seen up to 100x acceleration in model training, Sutton said.

“Compared to when we used CPUs, NVIDIA GPUs give us a very fast impact search loop,” he added. “It’s electrifying to work with something that can have an impact so quickly.”

From running just 10 trials, Featurespace researchers and data scientists can now run thousands of trials, increasing the statistical confidence in their results, allowing them to deploy only the best proven models.

Sutton said even a 1% increase in detection of fraud discovered using the deep learning model could save large companies $20 million a year.

Featurespace typically uses recurrent neural network architectures on transaction flow data. This model pipeline makes it possible to evaluate the new actions of an individual via the behavioral context drawn from their past actions.

Financial Fortifications for All

Featurespace’s deep learning models prevented all kinds of fraud, including those involving credit cards, payments, apps, and money laundering.

The ARIC Risk Hub interface is customizable, so clients can select the subset of components best suited to their specific needs. Users can then change scan settings or investigate suspicious cases. If, upon review, a case is deemed a false positive, the deep learning model learns from its mistakes, increasing future accuracy.

Featurespace technology has caused a stir with payment processors such as TSYS and Worldpay, as well as major banks such as Danske Bank, HSBC and NatWest.

As Sutton put it, “Featurespace uses AI to make the world a safer place to transact.”

“Our work is what brings many Featurespace people to the office every morning,” he said. “If you’re able to reduce the amount of money laundering in the world, for example, you can turn crime into something that doesn’t pay as much, making it a less profitable industry.”

Featurespace will host sessions on preventing fraud, money laundering and cryptocurrency at Silver 20/20a fintech conference taking place October 23-26 in Las Vegas.

Register for free for NVIDIA GTConline September 19-22, to learn about the latest technological advancements in the age of AI and the Metaverse.

To subscribe to NVIDIA Financial Services News.

Comments are closed.