VAST Data & NVIDIA Power AI Agents with NVIDIA Data Flywheel

VAST Data, known as the AI Operating System company, has unveiled a comprehensive platform that transforms how AI systems learn and evolve. By integrating the VAST AI OS with NVIDIA AI Enterprise, including NeMo microservices, the companies have developed a dynamic environment where AI can continuously refine itself. This synergy sets the stage for scalable, self-optimizing AI systems capable of evolving through the constant churn of data and interactions, ultimately enabling the emergence of billions of intelligent agents that improve over time.

A Unified Platform for Data, Compute, and AI Evolution

The joint solution delivers an all-in-one ecosystem that merges data management, database services, and orchestration of AI workloads. What sets the VAST AI OS apart is its AgentEngine, a powerful tool that captures and maps the complex relationships between agents and the data they interact with. Through detailed production logs, the platform gains granular insights into the learning processes of AI models. This traceability makes it possible to isolate and refine the specific factors impacting performance, dramatically boosting efficiency and accuracy across scaled AI operations.

“AI-powered businesses need thinking machines designed for a future where billions of AI agents learn from their own experiences, fine-tune in real time, and create new possibilities through collaboration,” said Jeff Denworth, Co-Founder of VAST Data. “By unifying NVIDIA’s AI software and hardware technologies within the core of the VAST AI Operating System, we are giving customers the foundation to operationalize continuous improvements in AI intelligence at scale, with the security, governance, and service delivery tools required to manage these intelligent agents and the data they rely upon.”

Real-Time, Autonomous AI for the Modern Enterprise

As AI becomes foundational to enterprise operations, there's a growing demand for systems that not only operate in real-time but also evolve with each interaction. AI is no longer confined to isolated experiments; it's transitioning into an always-on infrastructure. To support this shift, enterprises require platforms that adapt automatically to incoming data while addressing critical concerns like security, governance, and model fine-tuning. The partnership between VAST and NVIDIA enables the deployment of fully automated AI pipelines, spanning data ingestion, inference, and retraining, all governed by the VAST AI OS.

CACEIS Leads the Way in Real-Time AI Transformation

Among the early adopters of this transformative technology is CACEIS, a major asset servicing firm in Europe. In collaboration with VAST and NVIDIA, CACEIS is developing a real-time AI platform capable of capturing, transcribing, and analyzing every client meeting. The goal is to produce instant meeting summaries, uncover actionable insights, and offer anonymized trend analytics, all fed directly into a secure, sovereign CRM system. At its core, the platform is built on a robust security model, ensuring client data remains private while allowing the system to continually evolve through real-world use.

“AI will be a game-changer, highlighting trends in current needs by analysing meeting reports so we can better serve clients,” said Arnaud Misset, Chief Digital Officer, CACEIS.

Building the Next-Gen AI Factory at CACEIS

With the help of VAST’s AgentEngine that leverages the NVIDIA AI-Q Blueprint, CACEIS is testing a proof of concept for an intelligent platform that supports relationship managers during client engagements. This system, built around the NVIDIA data flywheel blueprint, is designed to learn from each interaction, refining its models in real time. The ultimate aim is to uncover new business opportunities by turning each conversation into a learning moment, where both human and AI agents learn from one another, fostering a collective intelligence that grows with every engagement.

Industry Momentum Showcased at GTC Europe

This powerful capability was demonstrated by Kari Briski during her presentation at NVIDIA’s GTC Europe event in Paris, highlighting how far enterprise AI has come in enabling real-time decision-making and continuous improvement. The showcase exemplified how cutting-edge collaboration can unlock smarter, faster AI ecosystems tailored to individual organizational needs.

Towards a Future of Self-Learning, Secure AI Ecosystems

This initiative underscores a larger industry shift: the move away from rigid, generic AI models toward adaptive ecosystems of agents that learn, collaborate, and innovate through their own experiences. Managing this new wave of AI requires not just technical sophistication but also strong security frameworks, governance models, and scalability. VAST’s AI Operating System rises to this challenge, offering real-time data infrastructure, intelligent workload orchestration, and policies that ensure fairness and safety, all within an adaptive, secure environment that empowers continuous AI advancement.

“Data flywheels leverage each interaction with an AI agent to continuously improve system intelligence and value,” said Kari Briski, Vice President of Generative AI Software at NVIDIA. “CACEIS is an exemplary pioneer with a vision of building an agentic AI data flywheel with NVIDIA and VAST to supercharge productivity for financial services in Europe.”