NVIDIA Blog· Jason Renders·· 9d agoSelected
European financial institutions adopt open models to build proprietary intelligence on NVIDIA accelerated computing
Original source: NVIDIA BlogRead original
AI brief
At GTC Berlin, financial institutions including Revolut and Starling Bank showcased production-scale AI built on open weight models and NVIDIA accelerated computing. Revolut is scaling its foundation model PRAGMA to service 80 million customers, while Starling Bank developed Starling Intelligence on U.K. sovereign cloud. Open models enable cost efficiency, with bunq's assistant Finn matching leading closed models at 60% cost per answer.
Why it matters
Readers interested in AI and finance should note how European institutions are leveraging open models and NVIDIA infrastructure to build proprietary AI capabilities while maintaining data control and reducing costs.
Source: NVIDIA Blog · blogs.nvidia.com