NAVER | D2 STARTUP FACTORY

In conventional Transformer models, the key-value cache (KV cache) grows with context length, limiting their ability to handle complex tasks on memory-constrained devices. TraceWeave addresses this challenge with Phase-TITAN, an alternative to the Transformer architecture, and proprietary AI models built on it. By keeping memory usage constant even as context grows, Phase-TITAN enables advanced AI agents to run entirely on-device, without relying on the cloud. Founded by researchers with backgrounds at ETH Zurich, Google, and KAIST, TraceWeave is targeting applications in mobility, robotics, drones, and smart glasses.

NXN Labs has developed a proprietary multimodal diffusion AI model that generates hyper-realistic fashion model images in under 30 seconds using only text and image inputs. It reduces time and costs by over 90% compared to traditional studio shoots while preserving the brand’s design identity. The platform serves paid clients including global luxury brands and major domestic and international fashion platforms such as KREAM and Charan.

We provide systematic digital education solutions to over 4,000 institutions, including elementary, middle, and high schools, universities, and various organizations through our AI-based educational solutions. Our proprietary sLM, specialized for education, creates AI tutor chatbots to enhance learning effectiveness, leading the digital transformation of public education. We also operate an AI cloud providing high-performance GPUs and NPUs essential for AI model training and inference.

Our SaaS platform enables users to efficiently create high-quality 3D assets using a proprietary 3D engine. These assets are compatible with various platforms, including Roblox and Zepeto. Our team includes experts who previously led service planning and 3D technology development at Naver, Samsung Electronics, and Kakao.

Rembrand developed In-Scene Media technology that seamlessly integrates brands into video content, offering a non-intrusive and engaging advertising experience for viewers. The advertising solution by Rembrand delivers eight times higher brand awareness and five times greater user engagement compared to traditional video ads. As a serial entrepreneurial team, Rembrand possesses deep expertise in Spatial AI technology, product development, and a vast network within the advertising business. Rembrand is founded by the team behind data platform BlueKai (acquired by Oracle) and AI platform Voicea (acquired by Cisco) as well as leading AI developers from Meta and Amazon.

Summer Robotics has developed a vision solution that leverages a laser event sensor capable of detecting light responses at the microsecond level, enabling robot control even in dynamic and unstructured environments. As a highly rare technology, it achieves four times faster speed compared to traditional robot vision sensors, ultra-precise measurement at the 0.1 mm level, and outstanding power and data efficiency. In particular, it demonstrates exceptional performance in automating tasks such as dynamic assembly, quality inspection, and component handling across manufacturing, logistics, automotive, and semiconductor industries. The co-founding team at Summer Robotics brings over 20 years of accumulated experience in technology research and product leadership in the fields of AI, machine vision, and robotics.

We are an AI company specialized in manufacturing sectors like automotive, semiconductors, and batteries. Our self-developed AI platform, "Runway," offers domain-specific AI and LLM for demand forecasting, production optimization, quality control, anomaly detection, and energy management. We have successfully applied over 5,000 AI models in the industry, leading the AX (AI Transformation) in manufacturing.

We developed a real-time markerless motion capture system utilizing a single LiDAR sensor and an AI model trained on proprietary datasets. This system accurately captures 3D motion for use in film, drama, gaming, and e-commerce for both live and virtual content. Our team consists of KAIST PhD graduates specializing in computer graphics and LiDAR software, with prior experience at Meta Reality Lab, NAVER Z ZEPETO, and SOS LAB.

We created a platform that connects over 400 shoe factories with more than 500 brands using data and AI. By reducing the time from product planning to production by up to 90%, our platform enables quick responses to trends and consumer needs while cutting manufacturing costs by up to 50%. We also offer GenAI solutions for shoe manufacturing, including trend analysis and design recommendations.

Our solution rapidly generates high-quality synthetic image and video data, facilitating the creation of training datasets that are challenging to collect in real-world settings. We are experiencing swift adoption across sectors such as autonomous driving, defense, smart agriculture, and smart homes. Our team possesses deep expertise in image processing and computer vision, with backgrounds from leading AI companies.