NamuAX Unveils Agentic AI Demo Based on NAA
IT DAILY ·
✦ AI Summary
NamuAX unveiled AI and cloud platforms and AI infrastructure operations technology at OpenTechNet Summit 2026.
At the event, NAA, SPERO, namuVIRT, Cocktail Cloud, and an NAA demo were introduced.
NamuAX unveiled NAAX and plans to strengthen its software capabilities for efficient GPU and AI model operations and expand into domestic AI infrastructure building.
NamuAX announced on the 18th that it participated in OpenTechNet Summit 2026 and showcased its AI and cloud platforms while also unveiling its AI infrastructure operations technology. NamuAX is the former Namu Tech. At the event, NamuAX CAIO Executive Director Go Woo-joo delivered a presentation.
At the event, NamuAX introduced NAA (NAMU Agentic AI), SPERO, namuVIRT, and Cocktail Cloud. NAA is an enterprise agentic AI platform, while SPERO is a multi- and hybrid-cloud management platform. namuVIRT is a server virtualization platform, and Cocktail Cloud is an AI-native platform.
NamuAX also demonstrated an NAA demo at the event. The demonstration showed how an AI agent carried out tasks using corporate documents and data.
NamuAX also unveiled NAAX, its AI infrastructure operations platform. NAAX is an AIOps platform that integrates and manages AI models and GPU resources for agent execution, and it is built on NAA and Xster, NamuAX's AI infrastructure operations platform. Through functions including GPU resource allocation connections, model deployment connections, endpoint creation connections, usage status management connections, and status management connections, NAAX links the processes needed to operate AI services.
Go Woo-joo, NamuAX CAIO Executive Director, gave a presentation on 'NAAX,' an AIOps platform for Efficient AI use of AI infrastructure. He said that as AI adoption expands, the key challenge is shifting from securing GPU resources themselves to allocating existing resources according to workload and operating a variety of AI models stably. He also pointed to project-by-project GPU overuse and manually centered model deployment as operational issues on site.
As countermeasures, he introduced MIG-based GPU partitioning and isolation, vLLM-based inference optimization, model catalog-based deployment, and endpoint creation. He said this approach enables workload-specific GPU partitioning and use, as well as standardization of model deployment procedures. He added that this approach is expected to increase utilization of limited AI infrastructure.
NamuAX plans to strengthen its software capabilities for efficient operation of GPU and AI models. It also plans to promote cooperation between its own platforms and Korea's AI server, accelerator, and data center ecosystems. It also intends to expand its business into the domestic AI infrastructure building area.
According to DealSite, NamuAX signed a memorandum of understanding (MOU) with financial data company Finova on the 7th to cooperate in the investment information content sector, and the scope of cooperation is AI-based investment information content business.
Source: IT DAILY · Kwon Young-seok
Original: https://www.itdaily.kr/news/articleView.html?idxno=241718
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.