In @Investment at that time (I decided to invest then), a working investor shares why he invested in this startup.
INEEJI, in which Capstone Partners invested twice, is a company with excellent sales capabilities. Jaesik Choi, CEO of INEEJI, started his career as a researcher at Somansa, Asia's No. 1 DLP company and a company specializing in personal information protection solutions. During his time at Seoul National University, parallel to his studies in computer engineering, his aptitude for technical sales seems to have developed from this point on.
Afterwards, he earned a doctorate in computer engineering from Illinois State University, worked as a postdoctoral researcher at Lawrence Berkeley Lab, and continued his research activities as an assistant professor in the Department of Electrical, Electronic and Computer Engineering at UNIST. He is said to have completed over 200 assignments to date, perhaps because of his nature to seek out work. As a hard-working professor, he experienced an inflection point in his career while working on the POSCO Smart Blast Furnace project.
◇POSCO Smart Blast Furnace’s good fortune and successive orders from large customers
While working at UNIST, CEO Jaesik Choi developed a model to predict blast furnace operation using a deep learning algorithm based on real-time data. Afterwards, he participated in the ‘Smart Blast Furnace’ project, which introduced artificial intelligence technology (AI) into blast furnace operations. To explain the iron making process of a blast furnace, that is, a blast furnace as we commonly know, in an extremely simplified way, iron ore and coke are stacked alternately in huge barrels, and through oxidation and reduction processes in which extremely hot oxygen at 1,200 degrees Celsius is blown, the iron ore is converted into pure molten iron. converted. At this time, it is most important to maintain the temperature inside the blast furnace at 1,500 degrees Celsius. Until now, skilled operators have been responsible for the quality of iron, the rice of the Korean industry, by directly controlling the intensity of hot air and pulverized coal input.
Currently, domestic steel companies are carrying out projects such as electric furnaces, melting furnaces, and cement kilns that are difficult to predict and control due to the complex and diverse process variables. It was not an easy path, but as we diligently carried out the task, INEEJI is building recognition among our customers as a representative AI company in the field of manufacturing process optimization. Among domestic AI research teams, they had the good fortune of having the rare experience of introducing AI solutions to large-scale heavy chemical processes, but it is said that the process of introducing AI solutions to chemical/heavy industry companies that are conservative and unfamiliar with the introduction of IT technology was not easy.
INEEJI categorizes its AI process optimization solution, ‘INFINITE OPTIMAL SERIES™’ into high-temperature reaction process, low-temperature/room-temperature reaction process, and failure diagnosis (predictive maintenance) solution and focuses on sales targeting industries in that field. did. As the introduction of the solution proved its effectiveness in improving quality, ensuring consistency, improving productivity, and reducing energy costs in the process, factories that were conservative in introducing software began to send PoC requests one after another.
The INEEJI team has accumulated know-how through projects with leading companies in each field in high-temperature reaction processes above 1,000 degrees and low-temperature and room temperature reaction processes below that. In particular, in the case of high-temperature reaction process control, it was introduced to the electric furnace process of steel companies, the heating furnace process in the continuous hot-dip galvanizing process, and the melting furnace in the glass manufacturing process, and in the low-temperature control process, it was introduced to the PO production process, etc., proving its effectiveness.
Based on the AI prediction and control know-how accumulated at large domestic heavy and chemical industries, INEEJI will turn its attention to the global market and make its first debut to Japanese customers at the Japan AI EXPO in October 2022, and then begin full-scale Japanese sales after establishing a Japanese branch in 2023. We are pursuing sales in Japan. After signing an MOU with GS Global Japan, a GS affiliated trading company, we are pursuing sales to win projects in the fields of steel, cement, oil refining, chemicals, and semiconductors along with a demonstration contract with a Japanese steel company.
It is not easy to do business with Japanese customers, who are much more conservative and closed-minded than those in Korea, but CEO Jaesik Choi and INEEJI employees, who never give up, are doing their best to move forward one step at a time without stopping. INEEJI recently attracted Series A investment worth 8.1 billion won from KT Investment, KDB Korea Development Bank, Woori Bank, and Capstone Partners and is preparing an RFP for major domestic listing managers for listing.
◇AI startup investment, technical skills, customer obsession and persistence learned through Ineasy
We are living in the era of the AI industrial revolution. Following the first AI shock caused by the AlphaGo/Lee Sedol match, we are experiencing the second AI shock caused by OpenAI's ChatGPT. Following the AlphaGo shock in 2016, deep learning AI startups with technological potential attracted a lot of investment, and following the ChatGPT shock in 2023, I feel a sense of déjà vu once again as I see investment funds pouring into generative AI startups.
The first investment in INEEJI was made in March 2022. ChatGPT's beta launch was in November 2022, so the initial investment was made at a time when AI was receiving the least attention in recent years. Over time, at the time of the second investment in December 2023, the popularity of generative AI startups peaked, but INEEJI's CNN-based time series prediction algorithm technology was treated as a technology that had passed the trend. Technology may have changed, but business may not have changed. The first customer must be successfully introduced to create a solution that can be sold to other customers, and the solution must be operated in the field 24 hours a day, 365 days a year to secure a data pipeline that can train a generative AI model.
The INEEJI team understands that customers commonly want a solution to predict raw material purchase prices, exchange rates, and demand for finished products, and provides a cloud solution that includes services that can identify factors that may affect future prices and establish strategies to prepare for price changes. It is being developed and is scheduled to be released within the year. The more difficult a technology, such as AI, is, the more difficult it is to introduce a solution in an industry with a low technology penetration rate. We decided to invest after seeing CEO Jaesik Choi and the INEEJI team's ceaseless efforts and persistence in customer sales, which are quickly filling the gap between technology and industry.
