It's after 4am. Hongjun An spent almost six months thinking about how to solve the image distortion problem using deep learning algorithms. After lying awake for hours, he jumped out of a bed. The idea of transforming the inputs in a different way just came to him while he was half asleep. He went straight to his desk and wrote it down in his notebook.

. . .

A Childhood with Physics

From a very young age, Hongjun actively sought out environments in which he could learn and grow. In primary school, he wanted to learn more about math and science. Without any help from his parents or teachers, he searched for the academy for gifted children and applied on his own. He also loved playing computer games and by the age of nine he had taught himself to code and was making his own games. At the school science fair, he was almost the only student to submit a robot in the competition and was chosen to represent the school.

When Hongjun was passionate about something, he didn't set limits, he dug deeper. His interest in robots led to electronic circuitry, and he borrowed books from the library and did a lot of reading. From the part where he reached the principles of semiconductors, he was introduced to the higher level of physics, such as that of electrons and electron holes. He realized that he would have to study physics in order to understand these properly. From then on, he was completely absorbed in physics. He continued to study physics in high school and university. Physics expressed the laws that govern the world in concise formulas. It was beautiful. 

It was during his high school years that Hongjun's grit began to show. His love of math and science led him to attend the Science High School for Gifted Students. Although his grades were good enough to graduate at the top of his class, the very first exam he took in high school placed him in the lower middle of the class. What he found most disturbing was not the disappointing grades, but that an invisible hierarchy had been created on the basis of grades. Driven by a sense of rebellion and defiance against a reality with which he couldn't agree, he pushed himself to study. He poured all his energy into his studies and was completely exhausted by the time he went to bed. And in just three months, he was ranked first in physics in the exam. Rather than being discouraged by the situation, he turned things around through extraordinary efforts. He had proved that an underdog could make it. And from that point on, he decided to become an inspiration to others.

More About the World Around Us: AI

With dreams of becoming a physicist, he enrolled in the physics department at Seoul National University. He developed his problem-solving skills by getting used to spending 10 hours pondering seemingly unsolvable physics problems. When he gained new knowledge, he wanted to share it with the world. Studying physics was great, but electrical engineering is about making an impact on the world in real time. He could imagine people benefiting from the technology he had developed. He decided to double major in electrical and computer engineering and started studying AI as an undergraduate intern in the lab of Professor Jongho Lee at Seoul National University. 

The more he studied AI, the more it suited him. His programming skills, which he had been honing since primary school, helped him improve quickly. Physics was also very helpful. The logical process of developing intuition from observation was applied to AI as well. On top of that, he devoted himself to his study in order to become better. More than a year after starting AI research as an undergraduate intern, he found a way to solve an MRI distortion problem that had been bothering him for about half a year. It was a deep learning solution that compensated for image distortion caused by breathing by measuring it with an artificial neural network without any additional equipment. The results were published in NeuroImage, the number one journal in the field of brain imaging. (‘DeepResp: Deep learning solution for respiration-induced B0 fluctuation artifacts in multi-slice GRE’) The potential of AI was developing for the benefit of the world. He was motivated to become a world-class researcher and reach the pinnacle of AI research. He went on to pursue his PhD, and in 2023, he was awarded the Summa Cum Laude Merit Award, which is given to the top 5% of the 6,000 candidates at ISMRM, the most prestigious international conference in MRI. (‘Active dictionary learning: Fast and adaptive parameter mapping for dynamic MRI’), Another great achievement as a result of 100% commitment.

Founding Wessist

“I know that whatever I do, I'm going to persevere and keep challenging myself to get it. Because I never give up, I'm going to grow. I just have to keep throwing myself into harsh environments. And I'll figure it out.”

Hongjun’s commitment to growth is reflected in his decision-making criteria: whether an experience is intense and whether he can grow from it. Having known him for 12 years since high school, Jeongjae and Eunji could see their potential unfolding if they teamed up with Hongjun. The trio, who were classmates at Gyeongnam Science High School, co-founded Wessist. They were all wary of complacency and willing to challenge themselves to grow. Crucially, they had the same problem-solving approach. When faced with a challenge, they didn't focus on the fact that there was no immediate solution, but instead empowered themselves to figure it out. 

After teaming up, they continued to testing out different ideas. As time went on, their determination only grew stronger. Hongjun never doubted that all of this was an element of growth because he believed in his team, and the company as an organism is capable of infinite development. As a startup founder, he enjoyed being pushed to work harder and became more and more driven.

Wessist's vision is an 'Optimized Life' for all by redefining efficiency and reshaping the way the world works with AI. Hongjun has been interested in tackling inefficiencies and helping others since he was young. In junior high, he developed an Excel program to help his mother managing her cram schoolwork and adopted OCR for textbooks. During his military service, he developed Excel macros to automate various tasks and built a service management system. As a graduate student, he took an outsourcing job developing a macro program to search and parse land registers. The time saved by the automation solutions allowed people to engage in more creative and value-added activities. Hongjun wanted to help people live more efficiently through AI. 

Wessist is currently focusing on the inefficiency of the marketing and advertising sector. Although social media marketing has expanded rapidly, most of the work of marketers is done manually. Wessist is developing and testing AI products that can automate and streamline the work of social media marketers. Technology should be accessible to all. Hongjun, as always, will persevere to the end and push the boundaries with the Wessist members who have the same goal in mind.