Partner content: How AI Factories Help Banks Keep Pace with Rapid AI Evolution

AI factories enable financial services firms to easily scale artificial intelligence

In 2025, AI budgets tripled to 20% of IT budgets. However, according to global IDC research, achieving ROI is still the biggest barrier to AI adoption.

This is why organisations need solutions that make it easier for them to quickly build and deploy AI models and applications, including AI agents, for a range of use cases, including businesses from the financial services sector.

In response, Lenovo and NVIDIA have partnered to deliver a validated, scalable, and enterprise ready architecture for AI factories.

An AI factory is a system of hardware and software designed to facilitate the creation, development, training, deployment, and management of AI models and applications. Lenovo’s AI factory is built on the Lenovo Hybrid AI Advantage Platform, powered by NVIDIA technologies, and is designed to accelerate innovation, simplify hybrid infrastructure, and reduce operational complexity. Together, these solutions meet the evolving demands of modern enterprises by strengthening productivity, performance, and security.

Delivering full-stack datacentre infrastructure that includes workstations, servers, networking, storage, partner solutions and services, the hybrid AI factory platforms are built with Lenovo Validated Designs based on NVIDIA Enterprise Reference Architectures and NVIDIA-Certified Lenovo ThinkSystem servers.

AI and financial services

The financial services industry is ripe for AI advances, with numerous use cases around managing risk, fraud detection and improving customer experience. Organisations have large volumes of data at their fingertips, but are subject to stringent regulations.

Speaking at Artefact’s Adopt AI Summit, AI for Finance talk in November 2025, Georgios Kolovos, Payments and Fintech Leader at NVIDIA shared that AI maturity in financial services has slowed compared to last year, as some organisations struggle to keep up with a rapid advances:

“We published a survey before the summer that showed that the maturity of banks when it comes to AI adoption has dropped compared to last year. And this is because the speed at which AI is developing. Last year, we’d barely started talking about Agentic AI. And banks traditionally are not set up to quickly respond to such dynamic environments.”

“This is where the concept of a full-stack implementation makes a lot of sense because the application layer is just the top. If you have your data and the control, you have the security, you have your compute power, changing the top layer is probably the easiest when the next evolution of agents or GenAI comes. But you need to have all of these elements in place.”

AI factories enable financial services organisations to scale the development and deployment of artificial intelligence more easily. Rather than running individual pilots, organisations can better manage the entire AI model lifecycle and integrate AI solutions including AI agents into workflows.

Kolovos explains that organisations can reap better insights from taking a full stack approach to AI factories:

“When we talk to customers and to partners, there's one very clear insight that comes from AI factories. Companies that are successful in deploying AI factories or AI in general are people that take a full stack approach.”

He highlighted that organisations need a strategy around application layer, how data is being used to generate insights, and a strategy around compute.

Partnering for AI success

Lenovo and NVIDIA are working with enterprises across industries to unlock AI’s potential. One organisation reaping the benefits is Deutsche Bank, which has an established partnership with NVIDIA to embed AI into its financial services operations.

Bernd Leukert, Chief Technology, Data and Innovation Officer and Member of the Management Board at Deutsche Bank said:

“We built our own digital assistant that allows our people to access any information and any document in the company. But as well, any public information…it's very relevant for bankers not to just rely on internal information, but also use external information.

“Then in phase two, we started by saying ‘we have existing processes, but these processes have too much manual work’. So we think that we should focus more on the customer service and less on back-office operations.

“We get thousands of unstructured documents every day via e-mail, PDF and Word documents. We have built an AI agent that is able to understand that text, the request from the customer, and automatically process it and trigger a workflow.”

Achieving AI ROI means going beyond individual pilots and industrialising AI delivery. For the financial services sector and beyond, AI factories have a key role to play in this, enabling algorithmic trading, enhanced fraud detection, simplified regulatory compliance and more.

Mauro Arruda EMEA Director, AI CoE at Lenovo explained that developing an organisational culture that support innovation is also key to success with AI factories:

“Financial organisations have a variety of data sources ranging from transaction to customer information and they’re all quite different. How do you build that and integrate that in the whole stack? And make sure the models and solutions drive value from this data.

“What we've noticed when we put the Hybrid AI Advantage framework to work with customers and internally, was that people are at the centre of this transformation. There's a cultural shift in how you integrate your employees and your workforce with AI agents. And we've integrated what we call adoption services and change management in the whole stack. There's no value from AI if your people are not using it, if they're not enabled to drive value from that technology.”

This article is sponsored by Lenovo.