Banco Santander, in conjunction with Nvidia, has rolled out new capabilities to evaluate AI models and agents for banking applications.
The Santander AI Lab, explores and tests emerging AI technologies before broader deployment across the bank. Further, it helps identify which model is best suited for each task and informs how to deploy it in the most efficient way.
Santander has been moving from large general-purpose models towards smaller, specialized models designed for specific banking tasks. In selected use cases, this approach has reduced serving costs significantly without impacting performance.
Through the Santander AI Lab, Santander has built Model Arena to evaluate open-weight models against banking tasks and identify the right model for each workload. Customer-service comparisons showed that smaller, specialized models can deliver strong performance on targeted banking tasks.
Santander has built specialized intelligence for core banking workloads on Nvidia Kumo, a family of open-weight foundation models for structured data. This allows companies to quickly turn structured data into predictive insights needed for agentic services.
“Nvidia is a key player in the infrastructure, software and research that is powering the latest advances in artificial intelligence. Combined with Santander’s scale across businesses, customers and geographies, we believe we are uniquely positioned to succeed in transforming those advances into practical banking capabilities,” Iñaki Bernal, chief artificial intelligence officer at Santander, said.
As part of the next phase of the initiative, Santander and Nvidia have established a three-year collaboration framework covering areas such as foundation models, AI agents, fraud and anti-money laundering capabilities, real-time decisioning and advanced simulation technologies, including a banking digital twin.
“Banking is one of the most data-centric industries. Data sits at the heart of every process and now banks need to become intelligence-centric. Transforming years of data into usable and proprietary intelligence is not about model size, but about training and fine tuning. In this new world, cost per intelligence output is a key metric. Santander is showing what that looks like in production with smaller models, specialised on the bank’s own data, evaluated against real banking tasks and served efficiently enough to run at the scale of a global bank,” Aser Blanco, global banking lead at Nvidia, said.