Config, a robotics data startup based in Seoul and San Jose, has secured $27 million in a seed funding round led by Samsung Venture Investment. The startup, which aims to build the data layer for robotics foundation models, received additional backing from Hyundai Motor’s venture arm ZER01NE Ventures, LG Technology Ventures, and SKT America, among others.
Heavyweight Backing
The funding round, which valued Config at over $200 million, also saw participation from angel investor Pieter Abbeel, co-founder of Covariant and a UC Berkeley professor, alongside financial backers such as Mirae Asset Ventures, Korea Development Bank, GS Futures, Kakao Ventures, and Z Ventures. This brings Config’s total funding to $35 million.
Founded in January 2025 by CEO Minjoon Seo, a former researcher at Meta and chief scientist at TwelveLabs, Config is focused on providing essential data for robotics learning and operation, rather than building robots themselves. Seo emphasizes the importance of quality data in enhancing robotic functionality.
Challenges in Robotics AI
Training robots differs significantly from training large language models, which rely on widely available internet text data. “Every piece of training data has to be physically collected,” Seo explained in an interview. This makes the development of robotics AI more expensive compared to software-only chatbots. As the demand for more capable robots grows, so does the cost of data gathering and labeling.
Config positions itself as the TSMC of robot data, providing essential data without competing with its clients. This strategy has attracted large manufacturers eager to develop proprietary robot AI instead of relying on external vendors.
Growing Clientele
Config is already generating revenue, according to COO and co-founder Jack Bang. Its clientele includes large manufacturers, system integrators, and companies in agriculture and defense sectors. Competitors in the space include Physical Intelligence, Generalist AI, and Skild AI.
The company records human motion data in controlled environments and in the field, operating facilities in Seoul and Hanoi with nearly 300 staff. To date, Config has collected over 100,000 hours of motion data, significantly more than AgiBot World’s 3,000 hours.
Future Plans
Seo outlined Config’s strategic priorities following the funding: scaling its data operations towards 1 million hours of collected data, growing its enterprise platform business to $10 million in ARR by the end of 2027, and launching a cloud-based robot-as-a-service product. This offering will allow companies to utilize Config’s foundation model without needing onboard hardware.
Config’s innovative approach to data conversion before training begins is described by Seo as akin to language translation, ensuring data is tailored for robotic application. “The data must be converted, not the model,” Seo stated, highlighting Config's core differentiator.
Source: https://techcrunch.com/2026/05/11/koreas-biggest-manufacturers-back-config-the-tsmc-of-robot-data/
