Enterprise-grade architecture for Agentic AI across FSI, combines The Universal AI Platform from Dataiku with NVIDIA accelerated computing
Dataiku, the Universal AI Platform has announced a new FSI Blueprint for deploying agentic AI systems in financial services. This blueprint is designed to help banking and insurance institutions create, connect, and control intelligent AI agents at scale, with the governance, performance, and flexibility required for production in these highly regulated industries.
This announcement builds on Dataiku’s integration in the NVIDIA Enterprise AI Factory validated design, which helps enterprises accelerate the development and deployment of secure, scalable AI infrastructure.
“AI agents represent the next major shift in enterprise productivity, and banks are among the earliest adopters,” said Malcolm deMayo, Vice President of Global Financial Services at NVIDIA. “This new bank blueprint from Dataiku, accelerated by NVIDIA, combines reusable components that enable banks to automate thousands of repetitive manual tasks. This allows institutions to deploy intelligent systems that can adapt to complex workflows and evolve responsibly over time—all while meeting regulatory and compliance requirements through central governance.”
The FSI Blueprint combines The Universal AI Platform and Dataiku LLM Mesh with NVIDIA NIMmicroservices, NVIDIA NeMo, and GPU-accelerated infrastructure. It leverages AI agents powered by NVIDIA to provide financial institutions with a secure and modular foundation for building agentic AI solutions across use cases like fraud detection, customer service, risk analysis, and operations automation.
“Financial institutions are under pressure to operationalize AI faster, while managing risk, regulation, and complexity,” said John McCambridge, Global Head of Financial Services at Dataiku. “This FSI Blueprint helps banks and insurers move beyond experimentation, delivering trusted AI agents that are observable, cost-controlled, and designed to deliver meaningful business value.”
The Dataiku LLM Mesh offers native integration with NVIDIA NIM to simplify deployment of open, proprietary, and custom models within financial environments. Guardrails within Dataiku LLM Guard Services, such as Cost Guard and Quality Guard, provide built-in oversight, giving IT and product teams control over model usage, cost optimization, and performance evaluation.
The collaboration between Dataiku and NVIDIA was unveiled during NVIDIA GTC Paris at VivaTech 2025. The FSI Blueprint represents the first in a series of joint initiatives to drive agentic AI innovation in highly regulated industries, with expansion planned into life sciences and energy.
Financial institutions interested in deploying the FSI Blueprint can engage directly with joint go-to-market teams from Dataiku and NVIDIA.
