Lee Min-hyung joined The Korea Times in 2014 and has worked as a journalist mainly in Korea’s finance, tech and automotive industry. He specializes in content creation, breaking news and in-depth analysis currently on transportation and mobility. You can reach him via mhlee@koreatimes.co.kr.
LG CNS launches agentic AI platform for large-scale system development
LG Sciencepark in southwestern Seoul, home to LG CNS' headquarters / Courtesy of LG CNS
By Lee Min-hyung
LG CNS unveiled an agentic artificial intelligence (AI)-powered development platform designed to automate the entire cycle of large-scale enterprise IT systems, the company said Monday.
This marks a significant step beyond the emerging trend of AI-assisted vibe coding.
While natural language-based coding tools have gained popularity for their ability to generate software code, they have largely been limited to coding assistance and often lack a deep understanding of enterprise system architectures and business contexts.
In heavily regulated industries such as finance, manufacturing and the public sector, development standards, security policies and legacy system structures further complicate AI adoption.
Against this backdrop, LG CNS launched the DevOn Agentic AI Native Development (AIND), a platform that deploys specialized AI agents to handle every stage of system development and operations, from analysis and architecture design to coding, testing and quality assurance.
When a user submits a request in natural language, specialized agents for such tasks in an end-to-end workflow deliver the final result.
For example, if a financial institution requests the development of an automatic transfer service linked to its account management system, the analysis and design agent first interprets the business requirements and designs the system architecture.
The coding agent then generates software code in accordance with the institution’s development standards. This approach enables users to focus primarily on reviewing and approving outputs, significantly reducing development time.
At the core of AIND is Knowledge Foundation, an ontology-based database that consolidates and structures enterprise IT assets, including development standards, security policies, source code and project documentation, into a format AI can understand.
The platform was jointly developed with Cline, a U.S.-based open-source AI coding company whose AI coding agent has emerged as one of the fastest-growing AI software projects on the global development platform GitHub.
The two companies plan to expand AIND deployments across the U.S., Japan and Southeast Asia, targeting security- and compliance-sensitive sectors, including finance, manufacturing, government and defense.
“By leveraging AI agents with expert-level understanding of enterprise systems, we will automate the development and operation of large-scale IT environments and drive productivity innovation for our clients,” said Ahn Hyun-jung, vice president and head of application architecture at LG CNS.