Amorepacific said on the 28th that it has built a research-specialized Generative AI platform based on research and development (R&D) assets accumulated over about 70 years.
Amorepacific integrated and standardized research data into a format usable by AI through the "Data Highway" project in collaboration with KT, and developed the research-specialized Generative AI service "AI Assistant LEMON" based on it.
Amorepacific established the nation's first cosmetics laboratory in 1954 and has accumulated research and development (R&D) assets such as raw material information, formulation data, research reports, patents, and papers. However, because the data were dispersed across multiple systems, there were limits to utilization, and to improve this, beginning last year it pursued the "Data Highway" project with KT, led by its in-house AX organization.
The two companies designed a platform based on "AI Ready Data" optimized for AI use and built an "AI Data Hub" that integrates millions of structured and unstructured data points. On top of this, they developed the research-specialized Generative AI service "AI Assistant LEMON," enabling research data to be searched, analyzed, and utilized across a variety of tasks, including not only research and development but also production, quality, and regulatory response.
In beta testing, gains in work efficiency were confirmed. In tests conducted with researchers at Amorepacific's R&I Center, the time required for research materials and regulatory review in the product planning stage was cut from about 12 days to 5 minutes. The time needed to review the research history of a specific raw material and its applicability to products also fell from about 6 days to around 5 minutes.
Seo Byeong-hwi, head of Amorepacific's R&I Center (CTO), said, "Good AI can become a true partner in research only when it meets good data," and added, "Data Highway and LEMON will be the starting point for creating a new research environment."