GTT unveils AI-based factory for AI networking and cloud-based security services

Fletcher Keister, chief product and technology officer, GTT, said its efforts began last year when it was looking at how it would provide services to its business customers based on what’s happening with AI.

“Our conversations with NVIDIA and Insight Enterprises were around what we could do with AI,” he said. “We decided to take a crawl, walk, run moving towards a sprint kind of approach because we knew something was going on here that we had to get to the forefront of, or we would get left behind pretty quickly.”

GTT’s initial step was to get network infrastructure to support the AI factory in three of its data centers: New York, London, and Prague.  

“As we started to build out this infrastructure, we realized we should build an AI architecture on top of it and do things in a way that gave us flexibility to not get locked into one thing or another because you don’t know what route it would take,” Keister said. “Everything is changing on a rapid scale, so the last thing you want to do is get locked into a decision but have a way to continue to move through and with the technology as it comes.”  

Four-legged partnership

As GTT looked for a flexible strategy that could adapt to changes, its partnership with Insight was key.

Besides being a preferred implementer for NVIDIA and Dell, Insight has a set of professional service capabilities. The company can unify platforms and embed AI where it delivers measurable impact. 

“This is how that four-legged partnership came together with NVIDIA as the internal AI brain,” Keister said. “One of the contributors to Dell’s success is that they have opened up a toolkit of tools to enable the building of agents, agentic workflows that sit on top of their GPUs.”

Dell’s Powered servers’ toolset, which they provide to enterprises, and the other solution components coming together have allowed GTT to move “at a faster pace than we would have.”

A replicable platform

With the AI foundation in place, GTT will be able to offer its AI platform in various locations that its network reaches.

The timing of this new platform comes as the service provider has continued to expand its network reach.

In May, GTT announced that it was expanding its EnvisionCORE global Tier 1 IP backbone in Latin America and the Asia Pacific. GTT is extending its network reach and performance in key metropolitan markets, in Argentina, Chile, China, Colombia, India, Malaysia, the Philippines, South Korea and Thailand.

Keister said that not only could the AI platform be offered in other markets, but it also gives it more elements to offer customers. “If you think about the replication of this platform from what we want to do from a customer-facing services portfolio,” he said. “It leads to our three-pronged strategy of building the AI-enabled enterprise for GTT to drive efficiency, differentiating the customer experience and innovating and creating new product capabilities of what we have in our AI factory.”   

One of the initiatives that GTT has put on its roadmap is to expand its Virtual Data Center (VDC) Infrastructure as a Service (IaaS) offering. VDC is an automated IaaS platform for cloud computing that allows businesses to design, provision, and manage their own cloud resources, including compute, storage, and network appliances, across 15 global zones in Europe and select locations in the U.S.

“We’re now reimagining this as part of a broader service offering,” Keister said. “As an extended set of cloud capabilities, we’re deploying infrastructure and compute to the customer premises in a private cloud environment and connecting that back to the public cloud.”

He added that this could be extended into customer environments. “Whether that’s a GPU on a managed edge device with virtual GPUs as part of a private cloud offering and how to extend into customers' use cases in the public cloud,” Keister said. “We would then look to bring that under a common management view and product innovation relative to this set of capabilities

Vertical market potential

While it is still early in the game, GTT sees that the initial demand for the capabilities that the AI Factory could be applied to many customer verticals.

In the near term, the service provider sees potential in the hospitality and manufacturing sectors.

For hospitality chains, the platform could be used to improve customer interactions and overall experience. Meanwhile, manufacturers could leverage the platform to integrate all of the input from all IoT devices and get insights into it closer to the source.

“What we’re finding in our conversations with customers is they are focused on how close to get the data to the models they are deploying,” Keister said.

He added that given the real-time developments of AI, GTT is “we are all learning as we go on this, but the early indicators are it will be more of a distributed platform and where those workloads are run, which aligns with our focus on multi-national companies.”

For multinational clients, the two biggest issues as more adopt AI are security of applications and data sovereignty, which is the concept that digital data is subject to the laws and regulations of the country or region where it is physically located.

During a recent event GTT took part in with its partners Insight and Dell in Spain, Keister said CIOs cited security and data sovereignty were their top concerns.

“Our focus is, and will continue to be, to position ourselves to solve their problems,” he said. “The conversation about AI is about moving together and coming up with broader experience, so we can be at the forefront of solving their problems rather than just launching technology because you can.”

For related articles, visit the Business Topic Center.
For more information on high-speed transmission systems and suppliers, visit the Lightwave Buyer’s Guide.
To stay abreast of fiber network deployments, subscribe to Lightwave’s Service Providers and Datacom/Data Center newsletters.