Consecutive Enterprise-Sector Wins Beyond Food Service: CommentMong Training Data Powers Public and Corporate AI Projects
Lemong has strengthened its position as a full-stack AI company by winning a contract with Korea Electric Power Corporation (KEPCO) to build an internal generative-AI–based standard system. Following its successful delivery of AI projects for the Korea Rural Community Corporation for two consecutive years, this KEPCO engagement marks another enterprise-sector milestone. The case is drawing attention as a rare example of a B2C-centric AI startup achieving sustained success in the enterprise market.
At the core of this project is Lemong’s proprietary AI framework, Luminir AI. Luminir AI consists of four key technologies: (1) customized LLM model development, (2) multi-agent architecture, (3) RAG (Retrieval-Augmented Generation) technology, and (4) unstructured-data intelligence. In particular, the RAG component—designed to learn from large volumes of internal documents so employees can instantly retrieve and generate the information they need—was applied to this project based on Lemong’s patented technology.
Lemong has already proven traction in food-service DX and AI. Through its AI-based review and comment management platform CommentMong, the company serves approximately 13,000 food-service business owners. Major franchise headquarters—including LOTTERIA, Goobne Chicken, and Pizza Hut—also use CommentMong Biz. The review data collected in real time from delivery and map apps, analyzed by AI and processed by multi-agent workflows to generate optimal context-specific responses, has accumulated into high-quality training data. This dataset is now recognized as applicable to public-sector and enterprise AI projects as well.
Recently, Lemong expanded its footprint in the food-service ecosystem by signing a strategic partnership with FoodTech Co., Ltd. for “MATE AI POS,” and also completed a successful POC with NICE Information Service. Co-CEOs Bo-hyung Kim and Hee-yong Lee stated, “Our goal is to enable anyone—from small business owners to public institutions—to leverage high-quality AI technology,” adding that the company plans to strengthen R&D capabilities by aggressively hiring specialized talent such as AI developers, data scientists, and MLOps engineers.
Meanwhile, Luminir AI has demonstrated international technical credibility, with a paper accepted at ICDM (IEEE International Conference on Data Mining), one of the world’s top three data-mining conferences. Related patents have also been fully registered, further reinforcing the framework’s reliability. The industry is paying close attention to how Lemong’s data and technology—accumulated through a food-service-focused B2C offering—are now expanding into public and enterprise AI engagements.
Lemong’s KEPCO contract is being evaluated as a representative case showing how a domestic AI startup can translate B2C-sector experience into successful public-sector projects, validating the pathway for broader entry into the enterprise AI market.
