Meet LG AI Research

Established in December 2020 as LG Group’s dedicated AI research hub, LG AI Research focuses on developing cutting-edge artificial intelligence technologies. The organization’s flagship achievement is EXAONE, which began as a 300-billion-parameter multimodal AI model capable of processing both language and visual data. Since its initial 2021 launch, EXAONE has evolved into version 4.0, featuring enhanced reasoning capabilities, expanded language support, and foundational elements for Agentic AI development.

Drawing from one of the world’s largest AI model training datasets, LG AI Research conducts fundamental research aimed at enabling more informed business decisions through advanced AI technologies.

To support its ambitious research goals, LG AI Research implemented NeuralMesh, WEKA’s high-performance storage platform, to streamline training of its foundational AI models in public cloud environments. With NeuralMesh, data bottlenecks between the organization’s GPUs and storage systems have been eliminated, enabling its multi-node AI models to scale more effectively, while significantly improving training throughput and enhancing the stability of its fundamental AI research operations.