AI Entrepreneurs Find Real-World Traction Through Korea’s TeX-Corps Program

SEOUL, Dec. 15 (Korea Bizwire) — Early-stage entrepreneurs using artificial intelligence to solve industrial problems are beginning to see tangible results, propelled by TeX-Corps (Tech EXploration Corps) — a commercialization program modeled after the U.S. National Science Foundation’s I-Corps that pushes researchers to validate their ideas through rapid, intensive customer discovery.

The program, designed to help AI researchers translate technological concepts into viable businesses, requires participants to conduct dozens of interviews in a short period, giving them practical insight into industries they aim to transform.

One standout is Snapscale, founded by 25-year-old POSTECH computer science student Kim Sang-yoon, who recently won the top prize in the “Chung Ju-yung Startup Competition.” Kim is developing AI tools to automate plant engineering design — a notoriously labor-heavy process in which small errors can take teams of engineers weeks to resolve.

Despite not having a chemical engineering background, Kim identified plant design as a major opportunity after interviewing 75 domestic and international engineers through TeX-Corps (Tech EXploration Corps).

The feedback helped him refine his focus from process simulation to full-plant design automation and confirm that much of the design workflow follows common patterns that AI could systematize.

Snapscale has since secured a proof-of-concept partnership with a 300-employee construction company, just one year into development. “Korea’s depth of know-how in plant engineering is world-class,” Kim said. “This is an area where Korea can lead the AX transition, and my goal is to become the ‘Nvidia of process engineering.’”

Another founder, Kim Geon-su (50) of Everyone’s Premium Transfer Fee, also credits TeX-Corps for turning an academic curiosity into a viable business. A former Citibank Singapore banker now studying technology management at KAIST, Kim began investigating why small business owners often face cash shortages even when their businesses are performing well.

He concluded that Korea’s opaque and poorly regulated system of commercial “premium fees” — large upfront payments required to take over an existing store — traps many entrepreneurs financially.

After interviewing more than 70 small business owners and real estate brokers through TeX-Corps, Kim realized the scale of the problem was far greater than he expected. He went on to develop an AI-based system to predict and manage premium fees, combining this with a fintech-enabled brokerage model.

His startup has since been recognized by Korea Asset Management Corp.’s TechBlaze program and KODIT’s NEST initiative.

A third example is Mindmile, founded by Lee Sun-mi (29), who long acted informally as a counselor to friends seeking mental health support. While researching depression and behavioral patterns at UNIST, she built a chatbot to help users avoid procrastination — a common early symptom.

But after conducting more than 200 interviews under TeX-Corps, Lee concluded that wellness apps rarely convert into sustainable businesses.

Instead, she pivoted to developing an integrated solution linking users, licensed counselors and counseling centers — offering AI-based intake assessments and guidance for those unsure about seeking help.

The concept resonated strongly with mental health centers, and Mindmile is preparing proof-of-concept trials. “My goal is to improve mental-health literacy,” Lee said. “If we can streamline counseling workflows, more people will get help faster.”

Together, the three founders reflect a growing trend: researchers and professionals using AI not for abstract experimentation but to solve entrenched, real-world problems — guided by a program that forces them to listen first, build later.

Kevin Lee (kevinlee@koreabizwire.com)