Provide foundation models with scalable, structured, multimodal training data that captures speech, vision, video, text, and real-world context.
Build the data layer for AI companies, turning raw global information into high-quality datasets for pre-training, fine-tuning, annotation, and automated processing.
SEMO = seeing hidden currents in data - modeling subtle signals, edge cases,
and transitions in complex systems.
2025.04.07
semo is accepted by Antler Japan and starts a 10-week Startup Program!
2025.01.15
semo gets approved by KERNEL Entrepreneur Community and joins KERNEL Hongo!
2024.12.14
semo APP participates in "Haichuang Cup" International Science and Technology Innovation Competition, Japan Division and achieves the 3rd place!
2024.11.27
semo APP participates in China (Shenzhen) Innovation & Entrepreneurship International Competition and achieves the 1st place among all seed projects!
2024.10.23
semo APP participates in The 2nd Innovation and Entrepreneurship Competition for Overseas Scholars, Asian Division!
2024.10.22
semo's CEO, Meishu SONG, was invited to pitch semo APP at The Global Women Entrepreneur Salon!
2024.10.31
semo APP participates in China (Shenzhen) Innovation & Entrepreneurship International Competition, Japan Division. We achieves the 3rd place and we will participate in the final competition! Here we come, Shenzhen!