As generative AI technology rapidly advances, interest in 'On-Device AI' is also growing. On-Device AI refers to AI technology that performs data computation directly on end devices, such as smartphones or tablets, without relying on central data centers or clouds. This technology can address the drawbacks of cloud-based AI, such as excessive power consumption, privacy risks, and latency issues. It is also known as 'Edge AI' because it operates on the edge of the network, without requiring an internet connection.

While cloud-based generative AI, like ChatGPT, brought significant attention to the AI industry in 2023, this year, On-Device AI is gaining attention as a solution that can effectively integrate AI into everyday life and workplaces.

AI industry experts predict that On-Device AI will be widely used in various industries, including autonomous vehicles, factory automation, smart homes and buildings, healthcare, and robotics and drones. According to global market research firm MarketsandMarkets, the global On-Device AI market size is expected to grow from $18.5 billion in 2023 to $173.9 billion by 2030, with an annual growth rate of 37.7%.

Major global tech companies and Korean giants like Samsung Electronics and LG Electronics are already competing in the development of On-Device AI technologies and products. In addition, Korean On-Device AI technology startups are rapidly expanding their businesses with recent investments.

On the 21st, B2B On-Device AI service startup 'ZETIC.ai' (CEO Yeonseok Kim) secured seed funding from Korea Investment Accelerator and The Ventures. The amount of the investment remains undisclosed.

Founded in April this year, ZETIC.ai has developed a serverless AI solution called 'ZETIC.Melange,' which allows AI models to run directly on mobile devices without expensive GPU servers.

With the Melange solution, developers can easily integrate AI models into their apps. Additionally, when clients provide their AI models and data, ZETIC.ai optimizes them and converts them into software libraries suitable for chips in each Android and Apple device. This process replaces all server operations with internal library operations, effectively eliminating server costs.

Moreover, ZETIC.Melange maximizes the performance of AI models by utilizing neural processing units (NPUs) of various System on Chips (SoCs). It optimizes not only Qualcomm chips, commonly used in Android devices, but also Apple's A-series and M-series chips. This not only reduces costs but also improves processing speeds on mobile devices and enhances data security.

ZETIC.ai plans to use the investment to further enhance its products and prepare for global market entry by hiring AI developers and other staff. CEO Yeonseok Kim stated, "For startups offering voice recognition AI services, cloud costs can amount to 10 million won per month or more than 100 million won annually. ZETIC.ai aims to eliminate service cost burdens by converting cloud AI services to On-Device AI services, creating an environment where AI services can be developed freely."

In early April, On-Device AI wearable camera solution startup 'Linkflow' (CEO Yongguk Kim) received 11.5 billion won in Series C funding from multiple investors, including SBI Investment. With this, Linkflow's total accumulated investment has exceeded 40 billion won.

Spun off from Samsung Electronics' in-house venture program C-Lab in 2016, Linkflow develops 'On-Device AI wearable camera solutions.' The company's On-Device solution is not only of high video content quality but also excellent in privacy protection, currently holding over 80% of the domestic public wearable camera market. Recently, it has expanded its reach beyond civil services to areas like the police and fire departments.

CEO Yongguk Kim said, "With this investment, we plan to launch new products targeting the B2C market this year and actively enter the U.S. market. We will strive to maintain competitiveness in various fields such as leisure, security, public safety, and defense."

Additionally, On-Device AI lightweight technology startup 'SqueezeBits' (CEO Hyeongjun Kim) received a 2.5 billion won pre-Series A investment from Kakao Ventures, Samsung Next, POSCO Technology Investment, and POSTECH Holdings at the end of January.