Yidu Tech Reports Continued Profitability in First Half of FY2025 with Sustained Positive Adjusted EBITDA

November 21, 2024, China's leading AI healthcare enterpriseYidu Tech (2158.HK)Announce the results for the six months ended September 30, 2024 (the “first half of fiscal year 2025” or the “Reporting Period”), and hold an earnings conference call on the following day.

During the reporting period, the Company continued to adhere to its “Focus” strategy, achieving total revenue of RMB 329 million. On a current business management basis, the Company’s adjusted EBITDA has remained profitable at over RMB 27 million since turning profitable in the previous fiscal year, with its interim profitability level basically on par with the full-year figure for fiscal year 2024. Benefiting from improved internal operational efficiency and enhanced synergies across business segments, the Company’s net loss for the period narrowed to RMB 56.36 million, representing a significant year-on-year reduction of 29.2%, while cash flow continued to optimize and improve.

At the press conference, Yidu Tech’s management introduced the “AI for ROI” strategy, committing to sustained investment in artificial intelligence technologies—particularly the research, development, and commercialization of its self-developed large language model tailored for the medical vertical. The company aims to expand the application scope of its core algorithm engine, YiduCore, thereby solidifying its industry position and competitive advantages. Management stated, “As a leading enterprise in China’s AI healthcare sector, the Company will fully leverage its extensive expertise in AI technology and healthcare, actively promote technology implementation and standardization, further drive the intelligent transformation of the healthcare industry, provide safer, more inclusive, and more precise intelligent solutions for the entire healthcare ecosystem, and create long-term value for investors.”

Data is a key determinant of algorithm performance. As of September 30, 2024, YiduCore, the core AI engine of Yidu Tech, had cumulatively processed and analyzed 5.5 billion authorized medical records. The medical knowledge graph derived from these records encompasses more than 100,000 medical entities, while the disease knowledge graph basically covers all known diseases. Coupled with multiple systematic, disease-specific datasets, this enables Yidu Tech to conduct more in-depth analysis and understanding of the complex characteristics of various diseases.

During the reporting period, leveraging a training corpus of over 500 billion meticulously cleaned and proportioned tokens, the company continued its research, development, and training of large language models (LLMs) specialized for the medical vertical. After comprehensively evaluating industry application costs and fundamental model capability requirements, the company focused on training a 70-billion-parameter (70B) model and successfully advanced the technological evolution of its medical AI agents.

In May this year, Yidu Tech ranked first among more than 100 participating models in the authoritative evaluation platform MedBench, launched by Shanghai AI Laboratory and Shanghai Digital Medical Innovation Center. On September 28, Yidu Tech’s large language model once again secured the top position in the large-model triage capability assessment organized by the National Health Commission.

In terms of scenario implementation, Yidu Tech’s large language models have been widely applied across multiple business scenarios, including healthcare, pharmaceuticals, and insurance. Specifically, the company leverages its AI middle platform to empower the medical sector with large model capabilities, enabling users to independently develop intelligent applications. Furthermore, it has launched a patient recruitment agent based on a large model for medical record understanding, thereby improving physicians’ efficiency in clinical trial recommendations and utilization of medical record data. Meanwhile, Yidu Tech provides physician empowerment tools that support the generation of popular science content and research insights.

During the reporting period, Yidu Tech continued to deepen its presence in its three core business segments, solidifying its market position, while simultaneously leveraging large language models to empower existing operations and foster integrated innovation.

During the reporting period, the Company launched a “Dual Middle-Platform” solution comprising a Data Middle Platform and an AI Middle Platform. In addition to providing data governance services, this solution offers specialized foundational large models tailored for the medical vertical and a comprehensive zero-code toolchain, thereby shortening the distance from data to intelligent application value. In the field of scientific research, empowerment by large language models (LLMs) has further consolidated the Company’s market position in clinical research services. By leveraging LLMs’ capabilities in medical record comprehension and structured extraction of key information, the efficiency of hospital-based scientific research has been significantly improved. Furthermore, the Tianjin Binhai New Area Economic-Technological Development Area Health and Medical Big Data Platform project, won during the reporting period with a contract value exceeding RMB 20 million, has also incorporated LLM-powered intelligent consultation services. The “Kaixin Health Tech” mini-program, a personal health advisor developed based on Yidu Tech’s proprietary large model, has been deeply integrated with the “Huimin Bao” (inclusive commercial health insurance) business. Leveraging advanced generative AI technologies and multimodal understanding capabilities, it provides exclusive services—including health records and health follow-ups—to over 5 million of the Company’s “Huimin Bao” users.

During the reporting period,Big Data Platform and Solutions SectorRevenue reached RMB 134 million. As of September 30, 2024, the company had provided solutions to 105 top-tier hospitals in China as well as 40 regulatory agencies and policymakers, covering more than 2,800 hospitals. The company also participated in over 20 national- and provincial/ministerial-level medical research projects.

Life Sciences Solutions SegmentRevenue reached RMB 145 million. During the reporting period, the number of life science clients served by the Company amounted to 71. The net revenue retention rate of the top 10 clients was 117.45%, and the average revenue per user (ARPU) increased by 57% year-on-year. Among the top 20 multinational pharmaceutical companies, 16 were clients of the Company. In terms of clinical research, the Company has completed 368 clinical studies, including sponsor-initiated clinical trials and investigator-initiated trials. The number of prospective and retrospective real-world study projects reached 261. As of September 30, 2024, the cumulative patient coverage exceeded 10,000 individuals.

Health Management Platform and Solutions SegmentRevenue of RMB 51.16 million was generated during the period. During the reporting period, the Company continued to deepen its engagement in the “Huiminbao” (inclusive health insurance) projects across core provinces and municipalities in China. It has served as the primary operating platform for the “Beijing Puhui Health Insurance” for four consecutive years, processing over 250,000 claims in total. The application of the expedited claims model has improved claims processing efficiency by 50%–70% compared to traditional methods. The Company also served as the primary operating platform for the “Shenzhen Huiminbao” for two consecutive years. Enrollment in the second phase exceeded 6 million participants, with an enrollment rate of over 35% and online enrollment accounting for more than 88%. As of September 30, 2024, the number of active users who had completed at least one transaction on the health management platform exceeded 24 million.

Looking ahead, Yidu Tech will continue to invest in and optimize its vertical large language models for healthcare, driving the deployment of its proprietary large models across various medical ecosystem scenarios. Meanwhile, the company will deepen its research and analysis of key disease areas to deliver more precise and efficient healthcare solutions. Furthermore, Yidu Tech is committed to enhancing customer experience by providing high-quality services, thereby creating greater value for industry participants within the healthcare ecosystem.