Chief Executive Officer Jang Seong-mook of SMHACCP spent 20 years at SK hynix overseeing planning and development of automation solutions, and came to the conclusion that "no industry is slower to undergo digital transformation than food manufacturing." He noted that of the 70,000 food manufacturing plants in total, 97% are small and midsize factories that still rely on manual work.

Based on that view, in 2020 Jang founded SMHACCP and has grown it into a Foodtech corporations that uses artificial intelligence (AI) to resolve inefficiencies on food manufacturing floors. The core is an AX (AI transformation) platform that makes manufacturing processes intelligent. It automates processes that used to rely on manual work, such as order intake, production, and shipment, to improve efficiency and accuracy.

Jang Seong-muk, CEO of SMHACCP/Courtesy of SMHACCP

Jang said, "Food manufacturing has more than 100 system menus, so only skilled workers can respond quickly," and added, "We are integrating them with a large language model (LLM) so that a single sentence in natural language can automatically issue instructions from production planning to shipment."

For example, if a client corporations says, "Please produce so it can ship tomorrow," the LLM automatically consolidation the order, production, and shipment systems to complete the work order. Through the LLM, precise work instructions and process control are possible.

Jang said, "SMHACCP's AX platform digitizes every step of food manufacturing." He added, "If manual work showed an accuracy level around 95%, an AI agent can raise it to 99%."

The company is also developing in-house functions for verifying compliance with the Food Labeling Act needed for new product launches, automatic label generation, automatic document completion, and root-cause analysis of process anomalies. Jang said, "Because of the nature of manufacturing, a general-purpose LLM can produce text that is completely different from the facts as if it were true," and added, "We are advancing our services with a Foodtech-specialized LLM based on accurate statutes and process data."

The company has been selected as a supplier for "Smart HACCP" and is working with the Korea Institute for Food Safety Management Accreditation. It is also preparing to expand its business under the brand "AIXERA." It is considering changing its corporate name from SMHACCP to AIXERA in the future.

SMHACCP work example/Courtesy of SMHACCP

An AI machine vision system that replaces visual inspection is also under development. In food manufacturing, there are dozens of product types, and every time the item changes, the equipment settings must be changed to catch errors. The problem is that changing equipment settings each time is not easy. To address this, the AI automatically detects defect types such as printing or quantity errors.

Jang said, "We are receiving positive reviews from food corporations and currently have more than 70 clients, including Dodram, Pulmuone, and Sempio," and noted, "We expect to post 4.1 billion won in revenue this year and reach 24.8 billion won by 2028."

SMHACCP offers a SaaS (Software as a Service) product priced at 300,000 to 500,000 won per month instead of an expensive on-premises system. Because small and midsize food manufacturing plants lack the capacity for upfront investment, the business structure was designed to make SaaS accessible. Another feature is recurring revenue, with existing customers accounting for 60% of total sales.

Jang said, "Food manufacturing faces hurdles to digital transformation due to high-mix, low-volume production, visual inspection, and manual management, but there is that much room for innovation," and added, "We will improve efficiency with AI and data and grow into a corporations with enough expertise that when people think of 'Foodtech LLM,' they think of us."

※ This article has been translated by AI. Share your feedback here.