[AI Revolution](126) AI Analyzes CCTV Footage... MayAI Presents Solutions for Offline Store Management

Interview with Park Junhyuk, CEO of Mayi
AI Solution 'Mesh' for Analyzing Offline Store Visitor Data

If you operate an offline store, you always have these questions in mind. Which products did the customers who visited the store today show interest in? How many people visited, how long did they stay, and how many of those visits led to purchases? Regardless of the products sold, these are core pieces of information that make up the store, but it is not easy to figure this out without asking each customer individually. Meiai is a startup that has stepped up to solve this problem. It does so through artificial intelligence (AI) technology. AI analyzes CCTV footage to inform how the store should be operated.

On the 18th, Park Junhyuk, CEO of Meiai, said, "Our goal is to make offline spaces smarter," adding, "Data-driven decision-making is essential for this, and Meiai is tackling the problem of ‘quantifying customer experience’ first." Founded by Park in 2019, Meiai offers an AI solution called ‘Mash’ that analyzes visitor data in offline stores.

Mash analyzes CCTV footage to identify visitors’ gender, age group, main movement paths, and even detailed routes of target customer groups. Through this, it can reveal reactions to specific areas within the store and interactions between staff and visitors. This data provides direction on how to organize the store to achieve key performance indicators (KPIs). It objectifies and automates customer information, monitors KPIs such as visit-to-purchase conversion rates and dwell time by store, and creates an environment where high-performing stores can be benchmarked. Park explained, "We apply our proprietary AI-based technology with deep learning algorithms optimized for analyzing visitors in CCTV footage, tracking their movement paths seamlessly," adding, "Our strength lies in accurately analyzing detailed data even in the low-resolution, special environment of CCTV."

Junhyuk Park, CEO of May I

Although visitors are filmed and analyzed, potential issues have been preemptively blocked. Mash recognizes all visitors based on full-body images rather than faces, pseudonymizing data to ensure safety from privacy violations or leaks. Another advantage is that it can analyze visitor data using existing CCTV without installing additional equipment. Based on this, Meiai currently counts major corporations with high demand for offline visitor analysis, such as Samsung Electronics, LG Electronics, Hyundai Motor Company, and CGV, among its clients. The company secured 6 billion KRW in Series A funding in the second half of last year and was selected as a ‘Baby Unicorn Company’ by the Ministry of SMEs and Startups in June this year.

Park, who majored in multimedia computing and machine learning at Yonsei University Graduate School, started the company with the hope that AI technology could solve real-world difficult problems related to images and videos. He said, "We analyze an average of 15,000 hours of footage daily in real-time with our AI engine, and since we began aggregating data at the end of 2020, we have analyzed over 8 million hours in total," adding, "Store managers can focus solely on strategic planning by utilizing data for improving store operation strategies, optimizing store layouts, and marketing."

Park believes that the parts focused on improving product quality, such as building good products, technology, and operational systems, have now reached a certain level of maturity. This is why they are concentrating on domestic market sales while actively planning global expansion. He said, "We think it’s best to first handle overseas bases of domestic clients to gain relevant experience and references," adding, "We are prioritizing expansion into Southeast Asia, which has cultural similarities to Korea, or Western regions where understanding and demand for data are higher."