"From Hours to 10 Seconds: Logistics Calculations Solved by AI"A KAIST Professor’s AI Startup
Entrepreneurial spirits are igniting a fresh breeze in the Korean economy as ambitious individuals dive into founding startups. Here are their stories.
In a future where fear and anticipation coexist, there’s a startup determined to accelerate a world where artificial intelligence (AI) goes beyond replacing humans to solving problems that humans cannot. That startup is "Omelet," led by Park Jin-kyu, an associate professor in the Department of Industrial and Systems Engineering at KAIST and an entrepreneur. Omelet is the developer of "Oasis," an AI combinatorial optimization solution that aids complex decision-making in industrial settings. We sat down with CEO Park to discuss the coexistence of humans and AI.
Park studied architectural engineering at Seoul National University, exploring structures that are both structurally sound and aesthetically pleasing. In 2009, he went to Stanford University in the U.S. for graduate studies, earning a master’s degree in electrical engineering and a PhD in civil engineering. Now, he is fully dedicated to developing AI solutions. “My majors may seem to have shifted slightly, but the core essence remains the same. The specific problems to solve may differ, but the answers lie in optimization and efficient operations. From architecture to artificial intelligence, I’ve simply been chasing interesting problems to tackle.
After graduating, he joined KAIST in 2016. For someone who enjoys problem-solving, the path of a researcher was a natural choice. “Engineering research, in particular, is tough to do alone. Teamwork is crucial. I loved discussing and solving problems with students at the university. For me, research and teaching were not separate—they were both methodologies for problem-solving. I wanted to find new and innovative approaches, and that desire grew into a passion for addressing the complex, real-world challenges industries face. That was reason enough to start a company.
His dream of entrepreneurship took shape in 2022 while spending a sabbatical at Stanford University. Coincidentally, his co-founder, Kwon Chang-hyun, an associate professor in KAIST’s Department of Industrial and Systems Engineering, also came to KAIST for a sabbatical that year. “A sabbatical is a golden opportunity to step away from routine work and pursue something new. I approached problems through artificial intelligence, while Professor Kwon tackled them with mathematical analysis. Our methods differed, but our goals aligned.
Based on their research, they developed "Oasis," an AI foundation model capable of solving complex problems. This approach starkly contrasts with traditional companies that merely adapt pre-existing algorithms into services. “It goes beyond simple office automation, like calculating the shortest route from Daejeon to Seoul. For instance, in delivery optimization, logistics companies can input their delivery list data, assign weights to goals like minimizing travel distance, reducing delivery time, or adhering to delivery schedules, and Oasis generates an optimized route based on those priorities. It’s incredibly user-friendly—just enter questions and conditions in a chatbot-style interface, and it solves the problem.
Currently, Oasis has entered the logistics market. “Beyond delivery optimization, it also optimizes the size of courier boxes. While traditional optimization servers took hours to compute when the number of delivery locations exceeded 1,000, Oasis can solve it in 10 seconds. It’s a fast, accurate, and novel way to tackle problems.