Nvidia is prominent for its footprint in the AI sector. In addition to its hot-selling AI chips, it is also helping European digital bank Bunq to stop financial fraud.

Nvidia Powers Bunq to Stop Financial Fraud

Bunq, dubbed as "the bank of the free," allows users to manage their finances entirely online via its app, eliminating the need for physical bank visits.

Since its inception in 2012, Bunq has grown significantly, serving over 12 million customers and managing deposits exceeding 8 billion euros. It is the first bank in more than three decades to receive a European banking license. 

With the help of Nvidia accelerated computing and generative AI, Bunq has enhanced the detection of fraudulent activities and money laundering.

According to the company, its automated transaction-monitoring system, powered by Nvidia's accelerated computing, has improved data processing and model training speed. The bank's AI-driven system leverages supervised and unsupervised learning to monitor transactions automatically and efficiently.

Thanks to the open-source Nvidia RAPIDS suite of GPU-accelerated data science libraries, Bunq has accelerated its data processing pipeline by over five times and trained its fraud-detection models nearly 100 times faster. 

Ali el Hassouni, the head of data and AI at Bunq, explained that they chose Nvidia's advanced, GPU-optimized software because it enables larger datasets and significantly speeds up the training of new models. That improves model accuracy and reduces false-positive rates, making the system more reliable and efficient.

Read Also: Nvidia Unveils AI Chatbot 'G-Assist' That Guides Gamers Through Games and Even Optimizes PCs

Personal AI Assistant of Bunq

Beyond its fraud detection efforts, bunq is integrating AI into various aspects of its operations. A prominent example of this initiative is Finn, Bunq's personal AI assistant, which aids users by providing insights into their spending patterns and transaction history. 

For example, Finn can report users' monthly grocery expenses or recall the name of a restaurant where they previously dined. Bunq is experimenting with Nvidia NeMo Retriever, a collection of generative AI microservices currently in early access, to improve Finn's functionality.

NeMo Retriever, part of the Nvidia NIM inference microservices, provides models as optimized containers through the Nvidia AI Enterprise platform.

El Hassouni highlighted Bunq's ongoing commitment to enhancing and broadening its AI applications for the benefit of its users. The partnership with Nvidia has been instrumental in advancing Bunq's technological capabilities, enabling the delivery of more efficient and precise services.

"Our initial testing of NeMo Retriever embedding NIM has been extremely positive, and our collaboration with NVIDIA on LLMs is poised to help us to take Finn to the next level and enhance customer experience," El Hassouni said in a press statement.

"We're creating a borderless banking experience for our users, always keeping them at the heart of everything we do," he added. 

Related Article: Microsoft Looks to Remedy Nvidia H100's Constricted Supply With Alternative AMD AI Chips