LG CNS launches AI agent development platform to automate enterprise IT systems

LG CNS has unveiled an AI-based development platform designed to automate the entire lifecycle of enterprise IT systems, from construction to operation.

The company on Monday launched Devon Agentic AIND, an agentic AI-based development platform.

"Vibe coding," a natural-language AI coding approach that has drawn growing attention, is widely seen as having limitations when applied to large-scale IT system development. The new platform addresses those shortcomings by deploying specialized AI agents for each stage of the process. When a user enters requirements in natural language, three dedicated AI agents work in concert to carry out the full development cycle: a customer requirements analysis and design agent, a coding agent, and a testing and quality verification agent.

For example, if a financial institution's user types, "Build an automatic transfer service for savings and time-deposit accounts linked to the account system," the platform analyzes relevant documents to design the system architecture. A coding agent then takes over that design and generates code in line with the institution's development standards. The company said users can focus on reviewing and approving results, significantly cutting overall development time.

LG CNS cited the platform's "Knowledge Foundation" — which integrates and analyzes the IT information needed for system development — as AIND's core competitive strength.

The Knowledge Foundation structures a company's IT information, including development standards, security regulations, system source code and development outputs, in a form that AI can understand. AIND uses that foundation to learn a client's systems and workflows, enabling customized development tailored to each customer.

LG CNS also applied a "spec-driven development" approach to AIND. Under this method, AI carries out design, coding and verification according to predefined standards, ensuring consistent quality regardless of the individual user.

AIND also supports "legacy modernization," converting existing systems into architectures optimized for the latest technology environments regardless of programming language. Tasks such as code analysis, conversion and verification that previously took weeks can now be completed in minutes, the company said.

LG CNS co-developed AIND with Cline, a US-based open-source AI coding company. The two companies plan to expand AIND's application to IT system construction and operation projects for global enterprises in finance, the public sector, manufacturing and defense, with a focus on the United States, Japan and Southeast Asia.

Ahn Hyun-jeong, executive vice president of application architecture at LG CNS, said the platform would "automate large-scale IT system construction and operation based on expert-level AI agents that understand enterprise systems, contributing to productivity innovation for corporate customers."

Meanwhile, LG CNS continues to strengthen its AI enterprise platform capabilities. The company also signed an "Agentic AI-based Process Intelligence Project Agreement" with Celonis, a German AI process development firm.