NAVER hosted its largest technology conference, DAN25, from November 6–7 at the COEX Grand Ballroom in Seoul. Over the course of two days, approximately 18,000 technology leaders, partners, developers, and general attendees gathered to explore NAVER’s latest achievements and future vision.
DAN, meaning “platform” in Korean, is NAVER’s flagship conference where the company shares its vision and explores fundamental questions about what platforms should be. Each year, the DAN series showcases how NAVER’s technology organizations are building connections and driving innovation forward.
This year’s DAN25 demonstrated how technology can deepen connections across people, industries, and everyday life through enhanced connectivity and expanded experiences. The venue reflected this theme, transforming into one vast network where presentations, displays, and interactive experiences created an organic technological ecosystem. Let’s explore the highlights.
Cloud interactive zones: Experiencing the AI value chain firsthand
One corner of DAN25 featured cloud interactive zones where visitors could explore NAVER Cloud’s technological journey at a glance. The exhibition’s central theme—the path toward sovereign AI—walked attendees step-by-step through the entire AI creation process, from data center infrastructure to cloud platforms, hyperscale AI models, and digital twin applications. Rather than overwhelming visitors with technical details, we focused on demonstrating how NAVER Cloud builds and operates comprehensive AI infrastructure independently, sharing both the technological foundations and underlying philosophy.
The data center zone showcased models of NAVER’s hyperscale facilities, Gak Chuncheon and Gak Sejong, highlighting Korea’s largest GPU clusters, uninterrupted power systems, and eco-friendly natural cooling design. An actual section of the building’s exterior wall demonstrated the energy-optimized cooling system and LEED Platinum–certified sustainable operations that captured visitors’ attention.
In the cloud platform zone, visitors explored NAVER Cloud Platform’s enterprise-tailored AI environment, featuring over 200 services alongside Neurocloud’s comprehensive security, integration, and flexibility capabilities. This demonstrated how unified infrastructure enables seamless AI model development and service operations from a single platform. Visitors also learned about HyperCLOVA X success stories—the hyperscale language model trained on Korean language and culture—and explored its various applications.
The recently released open-source HyperCLOVA X SEED drew particular attention, surpassing 300,000 downloads in just one month and catalyzing domestic AI ecosystem expansion. Real-world implementations—including CareCall, the Bank of Korea’s dedicated AI platform, and Korea Hydro & Nuclear Power’s generative AI system—illustrated how AI addresses challenges across social and industrial sectors.
The digital twin zone featured the smart city project in Saudi Arabia, demonstrating how cities are replicated in digital environments. Through videos and models, visitors saw simulation technologies for flood management, transportation, and urban planning supporting future city development decisions, highlighting the AI value chain’s expansion potential.
These interactive zones told a cohesive story of technological connectivity, spanning from data centers to city-scale solutions. They provided NAVER Cloud an opportunity to demonstrate the sovereign AI vision through tangible examples. Instead of diving deep into technical specifications, the focus was on highlighting underlying structures and principles, sharing the perspective on fostering AI ecosystem independence and diversity.
NAVER Cloud CEO’s message: “AI, a shared endeavor for all”
NAVER Cloud CEO Kim Yuwon delivered a keynote presentation at DAN25 titled “AI, a shared endeavor for all,” outlining NAVER’s technology philosophy and global vision for the AI sovereignty era. He emphasized that “AI shouldn’t be the exclusive domain of a few companies—it’s something everyone should participate in building together.” This message framed his vision for full-stack technological independence and AI ecosystem expansion that connects people, industries, and nations through sovereign AI. Here are the key insights from his presentation.
Sovereign AI: Self-directed AI
Sovereign AI means building an independent AI ecosystem that reflects each nation’s unique language, culture, and values using their own data and infrastructure—or “self-directed AI.”
“Creating an AI ecosystem that reflects one’s own language, culture and value system using one’s own data and infrastructure—that’s what we call sovereign AI.”
— Kim Yuwon, NAVER Cloud CEO
To realize this vision, NAVER has dedicated the past year to building a Korean-centric AI ecosystem. Beginning in 2023 with HyperCLOVA X—the world’s first hyperscale language model not centered on English—the portfolio has continuously expanded with the lightweight DASH model, the reasoning-enhanced Think model, and the open-source HyperCLOVA X Seed. HyperCLOVA X Seed has been particularly impactful, recording over 2 million downloads since its release and establishing a foundation for an open-source AI ecosystem where developers both domestically and internationally can contribute directly.
Advancing toward K-AI: Achieving technological independence
NAVER Cloud is participating in the Korean government’s “proprietary AI foundation model (K-AI)” initiative, developing capabilities as a national AI representative. HyperCLOVA X is evolving into an any-to-any omni-foundation model that extends beyond text to integrate audio, image, video, and mapping data, with an even more advanced model planned for release by year-end.
Building a robust AI ecosystem requires powerful infrastructure as its foundation. NAVER operates Gak Sejong, Asia’s largest hyperscale data center, and adds tens of thousands of GPUs annually to maintain Korea’s largest AI cluster. During the recent Asia-Pacific Economic Cooperation (APEC) summit, we announced plans to deploy an additional 60,000 GPUs, contributing to national-level AI infrastructure expansion.
NAVER is also collaborating with Samsung SDS on the national AI Computing Center project, opening GPU infrastructure to government institutions and companies through GPU-as-a-Service. These initiatives collectively advance NAVER’s vision of “sovereign AI for everyone,” supporting and expanding technological independence across the nation.
Physical AI: Expanding into the physical world
Physical AI represents a next-generation technology transforming industries worldwide, enabling machines to recognize and understand spaces, make autonomous decisions, and perform actions independently.
NAVER began researching robots and autonomous driving in 2016, becoming the first to transition to foundation model-based physical AI in 2021. According to NAVER LABS, physical AI is defined by three core characteristics:
- Real-world data: Information collected from hundreds of robots deployed across NAVER headquarters and data centers, operating in actual work environments
- Cloud platform: A single intelligence that integrates vast numbers of robots and data
- On-board AI: Lightweight models embedded into devices for fast, efficient decision-making
Together, these elements create an interconnected physical AI pipeline that continuously interacts and evolves.
Evolving toward industry-specific vertical AI
Physical AI is expanding beyond defense logistics into diverse sectors including agriculture, manufacturing, and services. A prime example is NAVER Cloud’s collaboration with Daedong, Korea’s first agricultural machinery manufacturer and a leader in future farming technology, to develop agriculture-specific AI agents.
This agricultural AI agent automatically generates farming records from voice conversations, provides crop- and region-specific information based on historical data, recommends optimal farming schedules, and alerts users to government agricultural support programs. NAVER’s advanced speech recognition technology accurately processes regional dialects, ensuring accessibility for older farmers. We plan to integrate Daedong’s comprehensive data—equipment operation logs, soil analysis, and pest information—with AI to create a complete smart agricultural ecosystem spanning from seeding to harvest.
This industry-focused approach exemplifies vertical AI, which integrates AI models, infrastructure, and services to address sector-specific challenges and demonstrates how sovereign AI translates into real-world applications. NAVER currently partners with major Korean enterprises including the Bank of Korea, KHNP, Daedong, Hanwha, HD Hyundai, LS, Lotte, and Hyundai Motor to accelerate AI innovation across industries.
“Collaboration in industrial sectors that represent Korea—like automotive, defense, and shipbuilding—is more important than ever. To realize vertical AI, partnerships between Korea’s industry leaders and NAVER, which has comprehensive full-stack sovereign AI capabilities, are crucial.”
— Kim Yuwon, NAVER Cloud CEO
Going global with K-AI
NAVER’s sovereign AI and physical AI are already driving real change worldwide, with Saudi Arabia, Thailand, and Japan at the center of these transformations.
1) Saudi Arabia: Digital twins for future cities
Saudi Arabia is pursuing national-level digital innovation through Saudi Vision 2030. Supporting this initiative, NAVER Cloud and Saudi’s National Housing Company established joint venture NAVER Innovation, which has deployed digital twin platforms across major cities to enhance disaster response and urban management efficiency.
A map-based super app that integrates Saudi Arabia’s terrain and local data is also being developed, combining NAVER Map and Place’s proven user experience with cutting-edge AI technology. This platform aims to set new standards for improving Saudi citizens’ quality of life.
2) Thailand: Revitalizing tourism with AI
NAVER Cloud is pioneering industry innovation through strategic partnerships. The collaboration with Thailand’s Siam.ai, a sovereign AI company, develops tourism AI agents designed to help Thailand’s core tourism industry recover from the COVID-19 pandemic. The system provides personalized travel recommendations, language support, and locally-tailored guidance based on visitor data, enhancing both tourist experiences and industry competitiveness.
3) Japan: Inclusive AI for a super-aging society
NAVER Cloud extends beyond industry efficiency to realize “inclusive AI” potential for underserved populations. In Japan, where a super-aging society faces care shortages and social isolation challenges, NAVER introduced AI CareCall in Izumo. This service provides regular check-ins for older adults, offering emotional connection and warmth through AI-powered conversations.
Inclusive AI: AccessTalk, speech support for children with autism
Beyond AI CareCall in Izumo, Japan, NAVER Cloud continues developing inclusive AI solutions for socially underserved communities. One such initiative is AccessTalk, a speech support project designed for children with autism who have difficulty speaking and their families.
For parents of children with communication challenges, even basic daily interactions can feel overwhelming. They spend their days searching for glimpses into their child’s inner world through eye contact and gestures, while children may become increasingly isolated as expressing thoughts and emotions grows more difficult. NAVER Cloud’s AI bridges this communication gap by providing parents with situation-specific conversation guides and offering children personalized picture and symbol card recommendations to facilitate mutual understanding.
Through AccessTalk, NAVER Cloud demonstrates that AI can recognize human emotions and serve as technology that restores relationships. These are the moments when technology brings warmth and connection to someone’s life.
NAVER Cloud’s vision: AI that connects all
Having advanced models and infrastructure to establish a sovereign AI foundation, NAVER Cloud is now entering the next phase of this journey—the era of AI agents and practical AI utilization. As CEO Kim Yuwon emphasized, “AI shouldn’t be the exclusive domain of a few companies—it’s something everyone should participate in building together.” To achieve this vision, NAVER Cloud is creating an environment where everyone can leverage AI in their field, regardless of technological expertise or financial resources.
Central to this new era is Neurocloud, the private cloud solution providing a secure environment for data and AI operations, with an enhanced version launching next June. The HyperCLOVA X model is also evolving beyond text understanding into an any-to-any omni-foundation model that seamlessly connects all formats—audio, image, and video. A comprehensive lineup spanning from large-scale models to lightweight physical AI solutions is under way.
Building on this foundation, NAVER Cloud is preparing to launch the “smart AI agent” project, designed to help small and medium enterprises, local businesses, and government offices easily access AI technology. The goal is to establish AI as a genuine partner across industries and regions—from agriculture and fisheries to convenience stores, community offices, and care services. This will enable farmers to forecast crop yields accurately, fishers to optimize operations using climate data, local governments to streamline administrative processes, and small businesses to deliver personalized customer management. AI will integrate naturally into daily life, becoming truly accessible to everyone.
The essence of AI lies in connection—technology should bridge industries and people, data and value, nations and communities. With comprehensive expertise spanning the entire AI value chain, NAVER Cloud is building a sustainable ecosystem that empowers governments, companies, and individuals to leverage AI solutions tailored to their unique needs.
Join NAVER Cloud’s journey as we create “AI for all” using our local data, infrastructure, and technology. Support our pursuit of sovereign AI that serves everyone, everywhere.