Leading South Korean manufacturing and technology companies are investing in Config, a fast-growing startup focused on building data systems for robotics artificial intelligence.
The Seoul- and San Jose-based company recently secured $27 million in an oversubscribed seed funding round led by Samsung Venture Investment. The investment pushed Config’s valuation beyond $200 million and increased its total funding to $35 million.
Several major strategic investors also joined the round, including ZER01NE Ventures, LG Technology Ventures, and SKT America.
The investment reflects Asia’s growing focus on “physical AI,” where artificial intelligence is applied to robotics, automation, and real-world machine interaction rather than only software and digital services.
Config was founded in 2025 by Minjoon Seo and a team with experience at major technology companies including Meta, Google, Waymo, and Naver.
Instead of building robots directly, the startup focuses on creating and improving the data used to train robotics AI systems. The company believes high-quality training data will become one of the most important parts of future robotics development.
According to Seo, developing robotics AI is much more difficult and expensive than training traditional AI chatbots because robot data must be collected through physical movement and real-world interaction. This requires robots, testing facilities, and human operators to generate usable training information.
Config compares its business model to TSMC, which produces chips for global technology companies without directly competing with them. Similarly, Config aims to become a core infrastructure provider for robotics AI companies by supplying specialized data instead of manufacturing robots itself.
The startup already works with manufacturers, system integrators, and businesses in sectors such as agriculture and defense. It currently operates large-scale data collection activities in Seoul and Hanoi, supported by a workforce of nearly 300 people.
To date, the company says it has collected more than 100,000 hours of human motion data used for robotics training, significantly larger than many existing open-source robotics datasets.
Unlike traditional robotics AI methods, which adapt human movement data after training, Config focuses on transforming the data before training begins so that it better matches how robots move and operate in real-world environments.
The newly raised funding will be used to expand operations, increase its robotics data library to one million hours, grow enterprise services, and launch a cloud-based robotics platform that allows businesses to run robotics AI systems without requiring expensive onboard hardware.
Industry analysts view the investment as another sign that Asia’s manufacturing-driven economies are becoming major players in the future of robotics and physical AI technologies.
Source: International technology and AI industry reports
