New stock outlook | Extreme perspective sprinting Hong Kong IPO: AI vision track "sell shovel people" potential can be expected, profit fluctuations dilemma difficult to solve.

In recent years, the cross-industry penetration of AI technology is accelerating the reshaping of industrial patterns, from precision medical diagnosis to intelligent industrial production, from real-time financial risk control to precision agriculture, its application scenarios continue to expand, and the depth of implementation continues to deepen. Among them, computer vision is one of the core and most mature sub-race tracks in the field of artificial intelligence (AI), with both technological foundation properties and industrial empowerment value. Recently, Shandong VisionDrill Technology Co., Ltd. (hereinafter referred to as "VisionDrill"), a Chinese AI computer vision solution provider, officially submitted an IPO application to the Hong Kong Stock Exchange, with CITIC SEC as its exclusive sponsor. It is worth noting that as a solution provider focusing on the field of computer vision, the IPO of VisionDrill is not only for its own development needs, but also reveals a "tip of the iceberg" of China's enterprise-level computer vision map. The "APP Store" of AI visual algorithms, with significant profit fluctuations The prospectus shows that VisionDrill focuses on providing AI computer vision solutions and large model solutions for enterprises: one is AI computer vision solutions. VisionDrill provides standard AI computer vision solutions, customized AI computer vision solutions, and software-defined integrated AI solutions. In addition, the company adjusts general large models to meet the diverse needs of customers by combining customers' industry and operating knowledge. By using advanced technologies such as multi-agent optimization, RAG, and scene-based algorithms, VisionDrill provides customized large model solutions designed for professional use in enterprises. Having been deeply rooted in this field for more than ten years, VisionDrill's core business focuses on the research and development, deployment, and full-process management services of AI computer vision solutions, providing end-to-end services to government and enterprise customers in over 100 industry scenarios including industry, energy, retail, and transportation. As of now, the company has built a comprehensive platform covering "cloud, edge, and end," creating an AI visual algorithm mall similar to the "APP Store," with over 1500 algorithms listed. As of the end of September 2025, it has established a global community of hundreds of thousands of AI algorithm developers and has provided services to over 3,000 customers, providing infrastructure platforms and AI solutions to help companies achieve digital transformation. During the period, the company has delivered over 6,000 projects, with a product repurchase rate of over 80.0%, demonstrating strong technical implementation capabilities. In addition, since the launch of large model solutions to the market in 2024, more than 80 entities have actively contacted the company to propose specific requirements for their business operations, reflecting strong market demand and the company's expansion potential in the field of large model solutions. According to data from Frost & Sullivan, based on 2024 revenues, the company ranks eighth in the emerging Chinese enterprise-level computer vision solution market, demonstrating a certain level of industry competitiveness. Financial data shows that VisionDrill has maintained a steady growth trend in revenue in recent years. In 2022 to 2024, the company achieved revenues of 102 million yuan, 128 million yuan, and 257 million yuan respectively. In the first three quarters of 2025, revenue further increased to 136 million yuan, a significant growth from 79.429 million yuan in the same period in 2024. The revenue growth is mainly due to the continuous expansion of the company's algorithm mall, the widening coverage of industries, and the gradual implementation of large model solutions. However, in contrast to the steady revenue growth, the company's profitability situation has fluctuated significantly. In 2022 and 2023, VisionDrill recorded losses of 60.722 million yuan and 56.246 million yuan respectively, mainly due to operating expenses related to cost structure and business expansion; in 2024, the company successfully turned its losses into profits, achieving a net profit of 8.708 million yuan, but the sustainability of profitability remains in doubt. In the first three quarters of 2025, the company's losses increased to 36.296 million yuan year-on-year. It is also worth noting that the company's operating cash flow has been consistently negative. From 2022 to the first three quarters of 2025, the net cash outflow from operating activities was 79.143 million yuan, 72.963 million yuan, 17.592 million yuan, and 21.373 million yuan respectively. Although the company has supplemented funds through financing activities, the continued operating cash flow gap may restrict its working capital, thereby affecting the stability of its financial condition. As of the end of September 2025, the company had cash on hand of approximately 19.684 million yuan. Fierce market competition, how to stand out? It is learned that computer vision is a typical sub-direction of AI (others include machine learning, natural language processing, Siasun Robot&Automation, cognitive computing, etc.). From the current industrial situation, the core technical difficulties in computer vision include algorithm design (challenges in robustness in complex scenes), hardware adaptation (contradiction and coordination optimization between algorithm efficiency and hardware resources), data acquisition and annotation, and various dimensions. Under the existing computational power conditions, optimizing technologies such as compilation optimization are key means to enhance algorithm efficiency and reduce computational costs. In terms of market prospects, data from Frost & Sullivan shows that the Chinese enterprise-level computer vision solution market is in a period of high-speed growth. It is expected that from 2024 to 2029, the overall market size will increase from 36.8 billion yuan to 182.4 billion yuan, with a compound annual growth rate of 37.7%; the growth rate in emerging markets is even more rapid, with the scale expected to soar from 11.1 billion yuan to 97 billion yuan, with a compound annual growth rate of 54.3%. The high-growth nature of the race track provides a broad space for VisionDrill's future development. However, from the perspective of market structure, the current Chinese emerging enterprise-level computer vision solution market is relatively fragmented, with VisionDrill's 1.6% market share significantly lagging behind the industry's leading company with 12.1% share, indicating a lack of a solid competitive barrier. As industry giants accelerate their layout and competitors in niche fields rise, it will be difficult for the company to increase its market share. To address this, VisionDrill's solution is to continue to increase research and development investment to maintain its technological matching with core capabilities. However, due to the fast iteration speed of technology in the field of computer vision, core technical difficulties such as algorithm design, hardware adaptation, data acquisition and annotation continue to exist, requiring companies to maintain a high level of research and development investment. From 2022 to the first three quarters of 2025, VisionDrill's R&D expenses were 35.2 million yuan, 36.6 million yuan, 44.8 million yuan, and 46.9 million yuan, accounting for 34.7%, 28.6%, 17.4%, and 34.4% of the revenue during the period, respectively. The continued growth in research and development investment provides support for technological innovation but also further increases the pressure on profitability. However, as a representative company in the AI computer vision race, VisionDrill, with its rich algorithm reserves, full-scene service capabilities, and industry first-mover advantage, is expected to seize more market share in the high-speed growth market. This push to the Hong Kong Stock Exchange, if successful, will provide important support for the company in terms of funding, optimizing financial structure, and enhancing brand influence. However, it should be noted that the lack of profitability stability, pressure on funding, and fierce market competition may all become performance drag factors after the company goes public. For investors, they need to closely monitor the company's technological research and development progress, commercialization efficiency, and changes in the industry competitive landscape.