WAVUS, a geospatial AI platform specialist, is joining forces with AI company Upstage to develop AI agents specialized in geospatial information and expand its presence in the public-sector AI market. By combining their respective expertise in geospatial information and AI, the two companies aim to build generative AI services tailored to the operational needs of public institutions.

WAVUS (Co-CEOs Hak-sung Kim and Chang-hoon Park) announced that it signed a memorandum of understanding (MOU) with Upstage (CEO Sung-hoon Kim) on October 8 for joint research and cooperation on AI agents.

Under the agreement, WAVUS will leverage Upstage’s large language models (LLMs) and AI foundation models to develop generative AI agent products powered by its GeoAI technology. Based on the technologies developed, the two companies plan to identify joint business opportunities and pursue projects in domestic and international markets. Specific joint business initiatives will be carried out under separate agreements.

The scope of cooperation includes joint ▲research and the application of LLMs in GeoAI ▲the use and joint validation of Upstage’s models in geospatial AI projects ▲the identification of domestic and international market opportunities and the pursuit of joint projects and ▲the development of generative AI agent technologies.

WAVUS has built up extensive technical expertise and project experience through public-sector initiatives in geospatial information, digital twins, and smart cities. More recently, the company has been expanding the application of AI agents in the public sector through initiatives involving urban safety AI, AI-based preliminary assessments for integrated permitting, digital twin technologies for national land management, and a GeoAI Agent project for the Korea Expressway Corporation.

Upstage is also expanding the application of its proprietary AI technologies in the public sector. Through the Ministry of Science and ICT’s Independent AI Foundation Model Project, the company is developing its own foundation models. It is also broadening the use of AI among local governments and public institutions through Solar Open 2 and Timely, its AI agent platform.

The geospatial AI agents the two companies aim to develop will go beyond simply processing location information such as addresses, coordinates, and cadastral parcel numbers. According to the companies, these agents will need to understand and utilize specialized information associated with geospatial data, including urban planning, permitting regulations, and administrative documents.

With the security requirements of public institutions in mind, the companies plan to develop geospatial AI agents that can operate in closed-network and on-premises environments. By combining general-purpose AI models with specialized geospatial data and domain-specific operational expertise, they aim to deliver practical services that can be deployed in real-world public-sector workflows.

“Geospatial information underpins public-sector operations in areas such as administration, urban planning, and safety, yet it remains a specialized domain that general-purpose AI models can struggle to handle effectively,” said Chang-hoon Park, Co-CEO of WAVUS. “By combining WAVUS’s experience in GeoAI Agent projects with Upstage’s foundation model technology, we will develop geospatial AI agents that public institutions can trust and put to practical use.”

“Geospatial information is a key domain in which Korean-developed AI models can demonstrate their value in real-world public-sector applications,” said Jae-yong Lee, Head of Public Sector Business at Upstage. “Together with WAVUS, we will develop proven GeoAI Agent use cases in the field and accelerate AI transformation across the public sector.”

[Access the original press release]