KT built a data platform that integrates Amorepacific’s research and development data and supports an AI-based research environment.

KT said on July 28 it completed the Data Highway project to integrate and redesign all data at Amorepacific’s Research and Innovation (R&I) Center.

KT converted millions of R&D records accumulated by Amorepacific over about 70 years into AI Ready Data that AI can understand and use after winning the project in December last year.

KT standardized structured and unstructured data that had been scattered across research labs and redesigned the structure so different data can be linked in a single context. The company said AI can understand and use information in an integrated way that researchers previously had to find separately across multiple systems.

The platform serves as a common data repository that continuously accumulates and manages research data and is not limited to a specific AI service. Amorepacific plans to use it to expand the application of various AI services and AI agents to R&D work.

The two companies also implemented the generative AI assistant Lemon (LEMON) to support researchers. When researchers ask questions in natural language, Lemon comprehensively analyzes data including raw material information, formulations, experiment results and research reports to provide the needed information.

For example, if a researcher asks about new raw material information for a specific formulation, the composition of fragrances and related experiment results, Lemon links the relevant data in a single context and presents the review results. It reduces time spent on repetitive information searches and supports researchers’ data-driven decision-making.

The two companies plan to continue expanding AI Ready Data and to upgrade the research environment so various AI services and AI agents can use it.

Noh Hyung-rae (노형래), an executive director and head of the Corporate Business Headquarters at KT’s Enterprise Division, said, "This is an AI transformation (AX) innovation case in which research assets accumulated by companies over decades were converted into data that AI can understand and use." He added, "We will expand AI-ready data-based AX across various industries such as manufacturing, distribution and services."