NTT DATA has launched a new initiative to help organizations deploy AI at scale. The company is now offering platforms powered by NVIDIA technology, designed to turn AI pilots into fully operational systems. These platforms combine NVIDIA’s GPU-based computing and fast networking with software tools to create a complete AI environment that works in the cloud and at the edge. The goal is to make AI development more straightforward and more reliable across different industries.

Building a Full-Stack AI Platform for Enterprises

The new platform integrates NVIDIA’s AI software, including NeMo and NIM Microservices, into a comprehensive AI stack. This setup supports every stage of AI projects, from training models to deploying enterprise applications. It operates within a governed framework, ensuring security and compliance. Abhijit Dubey, CEO of NTT DATA, explains that this approach reflects a shift in how businesses adopt AI. Instead of isolated experiments, companies now want ready-to-go solutions that deliver measurable results from the start.

NTT DATA believes their enterprise AI factory model bridges a common gap. Many AI projects struggle to move from successful pilots to full production. This platform aims to standardize results, cut down on the time needed to go live, and reduce costs. By providing a repeatable and scalable process, organizations can more confidently implement AI across various operations, from research to manufacturing.

Real-World Examples of AI Factory Deployments

Several organizations are already using these AI factories. One leading cancer research hospital leverages NVIDIA HGX platforms, along with NTT DATA and Dell, to improve radiology analysis. This setup helps researchers evaluate models quickly and supports clinical workflows more efficiently. In automotive manufacturing, a global supplier has shortened setup times by testing workloads on bare metal servers before scaling through the AI factory. This ensures smoother deployment and better resource management.

A third example involves a U.S.-based tech manufacturing company. They use NVIDIA-accelerated simulations and 3D visualization to test a new battery production line before building it physically. This approach saves time and reduces errors, making the process more cost-effective. These examples demonstrate how enterprise AI factories can be tailored to different sectors, providing sector-specific solutions built on a common NVIDIA infrastructure.

The Key Technologies Behind the AI Factory Model

The technical backbone of these platforms involves two main NVIDIA components. First, NeMo is a toolkit for creating AI systems that act like agents, built to run on GPU-accelerated hardware. It helps develop complex AI models that can understand and generate language, images, and more. Second, NIM Microservices offer ready-made, GPU-optimized containers with APIs. These enable quick deployment of AI applications in a variety of environments.

Together, NeMo and NIM Microservices form what NTT DATA describes as a full-stack, production-ready AI platform. They also provide pre-built generative AI prototypes, which help clients get started faster. This reduces the complexity of building AI solutions from scratch and accelerates the journey from proof-of-concept to operational systems. The combination of these technologies makes it easier for businesses to adopt AI in a reliable, scalable way.

John Fanelli, Vice President of Enterprise Software at NVIDIA, highlights the importance of this collaboration. He notes that enterprises are increasingly seeking platforms that can reliably scale their AI efforts from small pilots to full-blown production. NTT DATA’s AI factory offerings aim to meet this demand by providing sector-specific solutions built on a proven, flexible infrastructure. This move is designed to help organizations realize the full potential of AI across their operations.

Inspired by

• https://www.artificialintelligence-news.com/news/ntt-data-nvidia-enterprise-ai-factories-production/

Sources

• nttdata.com
• ai-expo.net
• techexevent.com
• techforge.pub
• techforge.pub