[AI Brief] Nutanix Launches “NAI 2.8”... Supports Agentic AI Operations in Production Environments
IT DAILY ·
✦ AI Summary
On the 27th, Nutanix announced the general availability of Nutanix Enterprise AI (NAI) 2.8 and said NKP 2.19 will also be generally available soon.
NAI 2.8 supports agentic AI operations in production environments, and Nutanix said it supports the simultaneous operation of legacy applications, data and AI through a governance-based dual-native architecture and integrations with major semiconductor partners.
Korea Deep Learning launched 'DeepAgent for Insurance,' a solution specialized for insurance claims processing, and Daim Research said it is cooperating with Intel to develop 'Davis,' an AI agent specialized for manufacturing and logistics.
On the 27th, Nutanix announced the general availability of Nutanix Enterprise AI (NAI) 2.8 and said Nutanix Kubernetes Platform (NKP) 2.19 will also be generally available soon.
NAI 2.8 supports agentic AI operations in production environments.
Nutanix cited the duality problem in enterprise AI adoption as the background for the solution launch. The company said AI adoption is accelerating the shift to containers.
It also said core applications and data are distributed across virtualized and containerized environments. As a result, enterprises must additionally bear separate infrastructure silos for existing applications and data as well as AI operations, along with the burden of moving data and redesigning existing workloads, the company said.
Nutanix said it proposed a governance-based dual-native architecture as a way to resolve issues that arise when running existing systems and AI together. The architecture supports the simultaneous operation of legacy applications and the latest AI, and Thomas Cornely, Nutanix's senior vice president and head of product management, said dual-native architecture makes it possible to apply AI to existing application and data environments.
Nutanix said this approach eliminates the need to redesign architectures by allowing AI to be introduced into environments where enterprise data already exists, rather than separating it into a new standalone environment. It added that it can be operated without network or data-layer complexity and helps reduce silos. In addition, it can leverage the most suitable infrastructure for each workload, and Thomas Cornely, Nutanix's senior vice president and head of product management, said it enables the use of optimal infrastructure by workload.
Nutanix said it has also expanded customer choice and flexibility through integrations with major semiconductor partners. Thomas Cornely, Nutanix's senior vice president and head of product management, said this makes it possible to implement consistent operations and governance across VM, containers and AI, and Nutanix said this also helps secure ROI.
(Photo: Korea Deep Learning)
On the 27th, Korea Deep Learning announced the launch of 'DeepAgent for Insurance,' a document AI solution specialized for insurance claims processing. The launch is intended to respond to the growing use of AI in the insurance industry.
AI use in the insurance industry had previously been limited to business support tasks such as counseling and search, but it is now expanding into core operations such as claims processing. Korea Deep Learning introduced the new solution in line with this trend.
'DeepAgent for Insurance' focuses on automating the stages after document recognition. The automation targets comparison, verification, and first-round review and decision-making.
The solution is used in insurance claim and review processes and supports more than 40 document types, including claim forms, medical certificates, medical expense receipts, detailed medical expense statements, and admission/discharge certificates. It has expanded its scope to include documents required for proxy claims, such as powers of attorney and family relation certificates, as well as various diagnosis and test documents.
It also includes a function that separates documents by claim conditions and a function that links and processes multiple documents as a single insurance claim. Korea Deep Learning said it designed the system to sort various documents needed for claims and reviews according to conditions and bundle them into a single claim for processing.
Korea Deep Learning conducted prior verification of the document processing performance for insurance claims before the product launch. Actual insurance claim documents and business criteria were used in the verification.
The processing environment consisted of integrated handling of a single claim involving multiple supporting documents, including claim forms, detailed medical expense statements, receipts, ID cards and bankbook copies. The items under review were document classification, data extraction, comparison and verification.
The verification results showed that work processing costs and time were reduced by up to 91% compared with conventional manual work.
Photo caption: Daim Research. Daim Research is developing 'Davis,' an AI agent specialized in manufacturing and logistics.
Daim Research is cooperating with Intel to develop an AI agent specialized for manufacturing and logistics sites. Daim Research was selected as a final participant in 'Ingenes,' an Intel-linked AI open innovation program under the Ministry of SMEs and Startups' global company collaboration program, and the selection announcement was made on the 27th.
The selection was made on the back of Intel's AI inference optimization technology. The development target is 'Davis,' an AI agent specialized in manufacturing and logistics, and Davis will run on-premises inside manufacturing sites without relying on external clouds, with a development focus on improving cost efficiency and security.
Daim Research has developed and supplied 'xMS,' an integrated management system for various logistics robots, and Davis will analyze the data accumulated in xMS using AI. Through this, Davis is expected to support the decision-making of on-site managers.
Site managers will be able to use natural-language queries to check how many robots are stopped or how a shutdown of a specific facility affects the production line, and responses will be based on actual operational data. It also provides functions that identify the cause of a failure and suggest countermeasures.
A Daim Research official said the rapid increase in robots and automation equipment has exposed the limits of the existing method of checking and judging all operational data, adding that the company is working to combine Intel AI technology with Daim Research's manufacturing and logistics data and robot control technology to address this. The official said the goal is to implement an industrial AI agent that can be used in actual manufacturing sites.
Source: IT DAILY · Yang Seung-gap
Original: https://www.itdaily.kr/news/articleView.html?idxno=241253
References
This article was produced with the help of an automated content generation algorithm.
Source: IT DAILY
View originalThis article was summarized and organized by BizCrush based on the original article from IT DAILY. For exact quotations and full details, please refer to the original article.