Highlights from NVIDIA GTC AI Conference 2025

Here's a recap of Capital One’s NVIDIA GTC sessions at NVIDIA GTC 2025, March 17-21 in San Jose, CA. Our team shared how Capital One leverages AI and data to innovate at scale through speaker sessions, booth presence and conversations throughout the conference with industry leaders and peers. 

Capital One's state-of-the-art multi-agentic conversational AI workflow was highlighted in NVIDIA CEO Jensen Huang's keynote, where he noted Capital One as one of the most advanced financial services companies in using technology.

Featuring: 

• Jensen Huang, CEO of NVIDIA

Watch the full keynote presentation.

Get insights on building an in-house, proprietary, multi-agentic conversational workflow for the enterprise. You’ll hear from Capital One about its journey to create a state-of-the-art and proprietary multi-agentic AI workflow that meets the needs of various users. The team discusses best practices, approaches and lessons learned from real-world experiments and applications for building an agentic workflow with several logical agents that understand natural language prompts, create an action plan, validate the plan and explain it back to the user before taking action on their behalf.

Featuring: 

• Milind Naphade, SVP, Head of AI Foundations, Capital One 

• Kamlesh Talreja, SVP, Financial Services Tech, Capital One

Watch the full speaker session.

Learn about how to strategize around, design, build and configure a high-performance computing (HPC) cluster with advanced GPU fault tolerances with Kubernetes on AWS. When running an HPC cluster, it's important to have robust recovery mechanisms to ensure minimal downtime and high availability. Take this to the next level by developing robust components from the ground up that integrate with Kubernetes to provide recovery from GPU XID errors in a seamless manner. We also cover best practices of building a scalable HPC cluster in a multi-tenant environment bundled with enhanced GPU data observability. 

Featuring: 

• Chandra Shekar Mudarpu, Director, AI/ML Platforms, Capital One

• Ali Rodell, Senior Director, AI/ML Platforms, Capital One

• Herik Webb, Director, AI/ML Platforms, Capital One

Watch the full speaker session.

Capital One is taking a well-managed approach to scaling enterprise AI in the regulated financial services industry. This fireside chat discussed Capital One's development and adoption of generative and agentic AI technologies, highlighting key initiatives that balance innovation in customer experience with well-governed and risk-centered approaches. Discover how the company creates differentiated value through proprietary, in-house AI development, including real-world examples of deploying multi-agentic systems that combine generative AI with rigorous validation frameworks. The discussion reveals critical lessons from implementing AI at scale—from technical execution challenges to fostering cross-functional alignment between practitioners and business leaders. Learn how Capital One's AI organization bridges cutting-edge AI research with practical applications, creating collaborative frameworks that transform technical capabilities into real business outcomes. 

Featuring: 

• Prem, Natarajan, PhD, EVP, Chief Scientist and Head of Enterprise AI, Capital One

• Jennifer St. John Foster, Senior Sales Manager, North America Financial Services, NVIDIA

Watch the full speaker session.

Get an in-depth look at how to accelerate distributed Spark applications using NVIDIA's RAPIDS framework on AWS EKS Kubernetes clusters. In this session, we present an overview of an analysis platform based on Kubeflow Notebooks open-source software, followed by an explanation of using Docker to build GPU-enabled Spark worker and scheduler node images. We walk through the setup of running Distributed Spark on Kubernetes and accelerating distributed Spark applications using RAPIDS and GPU-enabled workers, as well as the technical challenges of GPU-enabled distributed Spark applications and how to set up the RAPIDS Qualification Tool to profile our Spark jobs, and we share benchmarks and initial results for GPU-enabled distributed data engineering workloads. Preliminary results show significant cost savings for large GPU compute workloads. 

Featuring: 

• Bryan Nguyen, Lead Software Engineer, Capital One

• Tom Marthaler, Senior Manager Software Engineering, Capital One 

• Dhantha Gunarathna, Senior Software Engineer, Capital One

Watch the full speaker session.

Get a deep dive into how Capital One built a proprietary agent servicing tool for thousands of customer service agents across the business. Learn how the team used generative AI, advanced semantic search, and the NVIDIA Enterprise AI ecosystem to provide its agents with a natural, knowledge-grounded experience, and help them service customers effectively and efficiently. 

Featuring:

• Alfy Samuel, Director, AI Foundations, Capital One

• Tamara Sigler, Managing VP Servicing Strategy, Capital One

Watch the full speaker session.

New to tech at Capital One?

• Learn how we’re delivering value to millions of customers with proprietary AI solutions.

• See how we’re building and running serverless applications at a massive scale.

• Explore AI jobs and join our world-class team in accelerating AI research to change banking for good.