Decoding the core logic of medical data and AI, can RIMAG GROUP tell the story of Tempus AI in China well?

As the wave of artificial intelligence sweeps across various industries, the healthcare sector, due to its high barriers to entry and specialization, is seen as the most valuable and challenging "pearl on the crown" for AI applications. However, how to transform data into assets and AI technology into sustainable business revenue is a common challenge faced by global healthcare service companies. In the United States, TempusAI (TEM.US) has emerged as a darling of the capital market in the medical technology field with its unique "data + AI" model, reaching a peak stock price of $104.32 in October this year. Meanwhile, with the continuous deepening of the marketization mechanism for data elements in China, the exploration of the value of medical data is steadily advancing. In this context, is there an opportunity to create the next TempusAI in the Chinese market? Which companies have the potential to grow into platform-capable representatives of Chinese medical AI? Recently, the leading Chinese medical imaging service company RIMAG GROUP (02522) issued a data-related announcement. According to the announcement, in October 2025, RIMAG GROUP's subsidiary Beijing Yimai Guanghe Information Technology Co., Ltd. ("Beijing Yimai") signed a "Data Co-construction Cooperation Agreement" with Beijing Data Advanced District Service Co., Ltd. ("Beijing Advanced District") and Beijing International Computing Power Service Co., Ltd. ("Beijing Computing Power"). Cooperation will rely on the compliance regulatory framework of the data sandbox, leverage the medical imaging data resources provided by Beijing Yimai, with Beijing Advanced District responsible for data processing and compliance review, and Beijing Computing Power providing computing power and storage support, the three parties will jointly build a high-quality medical imaging data set, conduct large-scale model training, and explore innovative applications of medical imaging data. By the end of October 2025, the jointly built medical imaging data set had successfully achieved the first commercial contract within the data sandbox. The realization of this transaction reflects the gradual formation of a new type of collaboration model: In the "data sandbox" built by state-owned enterprises with clear rules, RIMAG GROUP, as a key provider of data services and technology, drives the entire value cycle, engaging in "value co-creation" of data. This model not only attempts to address the fundamental challenges and key obstacles of data circulation but also provides a reference for exploring a more suitable development path for medical AI in the Chinese market. It points more clearly towards a future: RIMAG GROUP is highly likely to become a medical data intelligence giant cultivated in Chinese soil that is even more valuable and imaginative than Tempus AI. Shared genes: Strategic upgrading from medical services to data platforms Both Tempus AI and RIMAG GROUP reflect a strategic transformation from traditional medical services to data-driven platforms. The core businesses of both companies fundamentally constitute a continuous, compliant, and high-quality data source. Tempus started with genetic testing and gradually integrated multiple sources of information such as clinical records, imaging data, and medication records in an attempt to build a patient's "digital twin." Its goal is to upgrade from a testing service provider to a provider of data infrastructure in the life sciences field. RIMAG GROUP's path is similar. By establishing and operating a third-party medical imaging center network in China, the company delves into the forefront of medical data generation, forming a positive interaction between business development and data accumulation. For investors, this "core business as a data source" model has a dual appeal. Firstly, it ensures the authenticity, continuity, and scalability of the data, which is a prerequisite for training reliable AI models. Secondly, the revenue from the core business provides stable cash flow to support the expensive data governance and AI output work, avoiding the financial risks of "burning money" in research and development for pure AI startups. Deepening data value: Anchor role of imaging data and multidimensional expansion potential Having data resources is just the starting point; the key lies in transforming them into high-value and tradable data products to build competitive advantages. In this process, the unique value of medical imaging data is worth exploring in-depth. Medical imaging data naturally plays an "anchor" role in personal medical information. Compared to the information gaps and non-structured data often found in clinical text data, imaging dataespecially in diseases like tumors that require long-term follow-upprovides objective, quantifiable information about lesion location, size, age, etc., serving as a stable and reliable reference baseline for clinical decisions. RIMAG GROUP has accumulated a certain level of standardization and refinement in the processing of imaging data. For example, the company initiated the establishment of the "Standardization of Medical Imaging Examination Items and Coding" early on, promoting internal data standardization and uniformity to lay the foundation for subsequent data circulation and platform applications. Leveraging its imaging center network and professional medical resources, the company has established multi-level quality control and annotation processes to gradually form high-quality and reusable annotated data sets. RIMAG GROUP's overlay driving mode of "standardized coding + fine annotation" has created its own characteristics in terms of data quality and usability. This model has some comparability in functionality to the barriers established by Tempus in the United States. Still, there are differences in achieving paths and applicable scenarios. It is worth noting that RIMAG GROUP has not stopped at imaging data but is extending towards multidimensional data. The company expects that the implementation of the Beichen project in Tianjin is driving its evolution from a single imaging data service to a comprehensive platform integrating imaging, pathology, laboratory tests (including genetics), physical examinations, electrophysiology, etc. This project covers five major modules: regional imaging sharing center, smart integrated laboratory, digital pathology center, health management center, and nuclear medicine center. This practice of integrating multi-modal data provides new possibilities for the deep value exploration of medical data. As the project progresses, if RIMAG GROUP can establish China's first large-scale, standardized, multi-modal medical data set, it may provide more possibilities for the company's business development and serve as a reference for improving the industry's data application level. From imaging data to multi-dimensional integration, the development path of medical data is constantly expanding, redefining its value space, and through the "Beichen model," RIMAG GROUP has taken an early advantage in this trend. Ecosystem construction vs. vertical integration: Financial mirror analysis of the business model of medical data platforms After the initial formation of data capabilities, RIMAG GROUP and Tempus AI show significant differences in their paths to commercialization. Tempus AI leans more towards vertical integration, focusing on serving pharmaceutical companies, insurance institutions, and other B-side clients, providing data licensing, analysis services, and SaaS solutions to support drug development and precision medicine, among other areas. RIMAG GROUP, on the other hand, prioritizes ecosystem construction, striving to create a "data-model-scenario" closed loop. On the one hand, through channels like data exchanges, it provides high-quality data products to AI research companies and research institutions, and on the other hand, in regional pilot programs like the "Yizhuang model," it participates in building data sandboxes to support compliant data collaboration. Additionally, the company is exploring the application of imaging data and AI capabilities in various scenarios such as improving primary healthcare, clinical trial evaluation for pharmaceutical companies, and analyzing the efficacy of medical devices. This difference in paths also reflects in their financial structures. Tempus' revenue is still primarily from laboratory services, with technology platform revenue gradually growing; RIMAG GROUP shows a "three-tier evolution" trend: the baseline level is still medical imaging center services, the second-tier high-growth engine is data empowerment and service revenue (including data trading, technology platform services), and as its data sets achieve successful transactions in data exchanges and the "data sandbox" model runs, this business line is transitioning from an exploratory "model validation" stage to a scalable "value realization" stage. This is a critical point where its valuation logic is about to undergo a qualitative change. The third tier envisions revenue from AI software and model licensing, which, with the maturity of its large models, could open up greater monetization opportunities through SaaS subscriptions, pay-per-use models, etc. In terms of business models, RIMAG GROUP's path leans more towards the positioning of a "ecosystem builder." It is not satisfied with just being a part of the industry chain but aims to become the core Hub Group, Inc. Class A of the entire medical imaging AI ecosystem by empowering upstream and downstream partners. The advantage of this model lies in its growth linked deeply with the growth of the entire medical AI industry in China, enjoying both the beta returns of the overall race and the alpha returns of the platform. It is worth noting that on October 30, RIMAG GROUP announced that it had made a nearly billion yuan capital injection into the core artificial intelligence company it incubated, Shanghai Yimai Zhiyi Intelligent Technology Co., Ltd., further increasing its equity stake. As a key AI company incubated by RIMAG GROUP, the core business of Yimai Zhiyi covers the operation of a comprehensive medical image intelligent service platform, AI-assisted diagnostic technology, and digital imaging solutions. Its technical research and product landing capabilities complement the existing imaging service ecosystem of RIMAG GROUP. The characteristics of "ecological complementarity and resource interconnection" have laid a foundation for deep cooperation between the two parties and have been a key consideration for RIMAG GROUP's increased investment this time. The long-term value of this investment lies not only in strengthening the company's strategic layout in AI technology research and product landing but also in realizing an organic connection in the business logic of "multi-dimensional data + AI layout + data trading," further solidifying the company's industry ecosystem barriers. Outlook: Development path and value considerations of data intelligence platforms RIMAG GROUP's recent investment in AI technology such as medical imaging large models reflects its intention to transform data value into intelligent applications. Based on large models, the company is expected to develop specialized AI applications at lower costs and higher efficiency, gradually building a model ecosystem for developers. The core logic is in constructing a "data-model-scenario" closed loop: starting from imaging networks and data governance capabilities, forming high-quality data products; leveraging large models to drive value creation; mature models then enhancing business efficiency. If this closed loop continues to operate, it will establish a self-reinforcing development cycle, becoming the company's long-term competitive strength. Looking at the industry prospects, in the wave where data elements and AI intelligence intersect, the "state-owned guarantee, data-driven, ecosystem co-building" model practiced by RIMAG GROUP provides another possibility for the development of Chinese medical data companies. Its value is not only in a single-point capability in technology or data but also in its ability to form a cross-scenario, sustainable intelligent ecosystem. This also requires the market to consider comprehensive assessments based on the higher-dimensional characteristics of the "data intelligence platform ecosystem" for its value evaluation.