![[엘리스그룹 사진자료] 엘리스그룹, 중기부 ‘2025년 스케일업 팁스 R&D’ 최종 선정.jpg](/_next/image?url=https%3A%2F%2Fcdn-cms.elice.io%2Felice-strapi%2F2025_R_and_D_c826213a93.jpg%3Fsp%3Dracwd%26st%3D2022-07-11T09%253A43%253A29Z%26se%3D2032-07-11T17%253A43%253A29Z%26spr%3Dhttps%26sv%3D2021-06-08%26sr%3Dc%26sig%3DXE8S5Wrx%252F6RCT1tL2PGNsACU1NvYYJAo7lDuH65zn1A%253D%26width%3D2000%26height%3D1125&w=3840&q=75)
KRW 1.2 billion in R&D funding over three years to accelerate development of high-reliability AI document processing technology specialized for manufacturing
Elice Inc., a full-stack AI company providing AI infrastructure and cloud, AI models, and industry-specific solutions, has been selected for the “2025 Scale-up TIPS R&D” program organized by the Ministry of SMEs and Startups (MSS). With this selection, Elice plans to accelerate the advancement of AI technologies tailored for the manufacturing sector by leveraging the secured R&D and commercialization funding.
The Scale-up TIPS program, overseen by the Ministry of SMEs and Startups and operated by the Korea Technology and Information Promotion Agency for SMEs (TIPA), is a technology growth program in which the government and private investors jointly discover and foster promising, technology-intensive startups and SMEs entering a full-scale growth phase. The general R&D type, for which Elice was selected, is a core support category focused on national strategic technologies, global AI competitiveness, and innovative AI companies.
Targeting the Asian manufacturing market with full-stack AI capabilities from infrastructure and models to industry-specific solutions
Elice was highly recognized for its full-stack competitiveness, which includes AI infrastructure design capabilities optimized for industrial environments, stable private cloud operation, and the in-house development and operation of AI models and solutions tailored to specific industry challenges. In particular, the company’s specialized AI model “Helpy Struct,” which can precisely extract and structure complex tables, drawings, and document data, along with multiple proprietary specialized AI models and solutions, were evaluated highly in terms of technological strength and business potential, leading to its selection for the Scale-up TIPS program.
Through this selection, Elice will receive a total of KRW 1.2 billion in R&D support over the next three years. The company plans to develop a high-reliability intelligent document processing (IDP) technology for the manufacturing industry that integrates Vision-Language Models (VLM) and multi-agent models on top of a private AI infrastructure. This technology will implement the entire process—from document collection and interpretation to verification and system integration—within a single AI pipeline, addressing productivity loss issues in manufacturing environments.
Elice aims to reduce misreading and misapplication of complex manufacturing documents with a single AI model, while implementing multi-layer AI agent cross-verification mechanisms to control AI hallucinations and achieve over 99% reliability. In addition, the company seeks to enhance versatility through zero-shot capabilities that enable immediate processing of new and complex manufacturing document formats without additional training.
With this project, Elice plans not only to accelerate the AI transformation of Korea’s manufacturing industry but also to actively expand into global markets. The company will focus on strengthening document processing performance for languages where general-purpose LLMs are relatively weak—such as Japanese and various Southeast Asian languages—in order to secure a competitive edge in the Asian manufacturing market.
Jae Won Kim, CEO of Elice Inc., said, “This selection is a recognition of Elice’s AI model development capabilities, security framework, and our strengths in private cloud and high-reliability AI infrastructure, which are essential for specialized industrial AX. We will continue to develop AI technologies optimized for the needs of manufacturing and other industries, capable of precisely supporting complex decision-making, and aim to become a trusted standard AI solutions provider in industrial sites around the globe.”
