Cognizant Beefs Up AI Portfolio With Neuro Edge Platform

‘You’re getting improved data privacy and security through localized computing, as well as reduced cloud dependency because now you're processing a lot of the data locally,’ says Vibha Rustagi, Cognizant’s global head of IoT and engineering.

Global IT services provider Cognizant Monday unveiled Cognizant Neuro Edge, a new addition to its Neuro suite of technologies aimed at helping businesses take advantage of AI and generative AI.

Cognizant Neuro Edge brings AI and GenAI to enterprise businesses, especially manufacturers, working with edge AI from chips and devices to applications, said Vibha Rustagi, global head of IoT and engineering at Teaneck, N.J.-based Cognizant, which is ranked No. 8 on CRN’s 2024 Solution Provider 500.

Cognizant Neuro Edge is a generic framework aimed at accelerating the development of enterprise edge services, Rustagi (pictured) told CRN.

“This can be done by abstracting the complexity of model selection by quantization, by fine-tuning of AI and GenAI models, and by RAG or retrieval augmented generation and so on agents that are compatible with edge devices,” she said. “We’ve introduced edge compute which means faster accessibility to data and analysis of the data, and then real-time analysis of that.”

The key importance in Cognizant Neuro Edge lies in the fact that businesses require hybrid GenAI cloud capabilities, Rustagi said.

“It gives faster access to data, which means businesses can leverage the available compute resources in edge devices,” she said. “And you're getting improved data privacy and security through localized computing, as well as reduced cloud dependency because now you're processing a lot of the data locally. So it really provides high-efficiency, low-bandwidth scenarios with much reduced lag and latency because of the real-time decision making happening on the edge. Combined with all that, and personalization and content contextualization, that's a huge advantage we'll see across the board with the adoption of GenAI on the edge.”

Neuro Edge is part of a wider Cognizant Neuro platform for simplifying and accelerating the adoption of automation and AI in the enterprise, in this case specific to the edge, Rustagi said.

It is an AI accelerator for any device or platform providers with a big focus on edge devices, such as cars, medical devices, surveillance cameras, or manufacturing equipment, Rustagi said.

“For example, for a device manufacturer or a manufacturing plant that's using a device, Neuro Edge could analyze the data for that specific manufacturing equipment,” she said. “We could provide predictive maintenance or similar results.”

While Cognizant has talked with companies in multiple industries, the company so far has only listed Qualcomm as a named customer, Rustagi said.

“We are working with Qualcomm to extend the Snapdragon Digital Chassis' generative AI capabilities,” she said. “What we haven't announced yet is, we're working with other customers and other industries, such as medtech and manufacturing, industrial IoT and in the telecom space. With Qualcomm, we are working to deploy GenAI at the edge to create new opportunities for automakers to develop highly personalized and contextually relevant experiences for both drivers and passengers.”

Cognizant Neuro Edge lets car manufacturers provide real-time personalized contextual awareness, as opposed to just generic information, and enabling new features in cars that may not have been previously possible, Rustagi said.

“It's also giving the OEMs in that specific example the opportunity to unlock many new revenue streams that were not possible earlier, and in general provide a lot of possibilities with industrial use cases that can be defined by the customer and the OEM in this specific case,” she said.

Going forward, Cognizant will expand the use cases of Cognizant Neuro Edge via new market segments, Rustagi said.

“We are working with some medical device providers or with industrial IoT providers to expand the applications of this to their use cases,” she said. “There are different hardware platforms we’ll work with, expand to various LLM 's out there in the market, expand further to specific use cases, and expand it with very unique data sets in these industry segments.”