South Korea-based semiconductor AI startup SemiAI has raised seed funding from Kakao Ventures. The investment amount was not disclosed.

The startup is building SMILE (Semiconductor Manufacturing Intelligence), an AI-powered software platform designed to boost semiconductor yield. Its core innovation lies in Virtual Fab Data generation—a technology that simulates and validates sensor, equipment, and wafer data across every stage of chip manufacturing. Engineers can use this virtual environment to spot defects across past, present, and future production runs at a glance, while uncovering the causal relationships between process variables that drive yield loss.

SMILE features Agentic AI that continuously learns from Virtual Fab Data in real time, automatically classifying defect patterns and streamlining yield analysis. The system studies engineers’ historical recipe adjustments and process optimization strategies to deliver tailored recommendations for each manufacturing step. According to the company, this approach cuts yield analysis and improvement cycles from over seven days to under ten minutes.

Armed with fresh capital, SemiAI plans to forge partnerships with global semiconductor manufacturers and fabs. The company will run proof-of-concept projects to validate its virtual data against real fab datasets, building credibility and demonstrating real-world applicability. The ultimate goal is to evolve into an AI infrastructure company that maximizes manufacturing efficiency by harnessing the massive data streams generated throughout semiconductor production.

Leading the company is CEO Taekwon Ji, who earned his Ph.D. in mechanical engineering from UC Berkeley before building an extensive career at industry giants including Intel, Lam Research, ASML, Samsung Electronics, and SK hynix. His team combines deep technical expertise with hands-on semiconductor industry experience, positioning them to drive meaningful change in how chips are manufactured.

“As AI computing power expands, memory efficiency will take center stage next year, making semiconductors more critical than ever,” said Youngmoo Kim, Partner at Kakao Ventures. “SemiAI has the execution capabilities to seize this moment and transform yield improvement—the fundamental goal of semiconductor manufacturing.”

“Fabless design for application semiconductors is important, but the industry desperately needs innovation in yield improvement to achieve mass production at scale,” said Taekwon Ji, CEO of SemiAI. “We’re building intelligent, automated systems that will lead the next wave of data-driven semiconductor manufacturing.”