The "Null Hypothesis" of AI Medical Productivity Tool Development Company recently received nearly 100 million yuan in Series A financing, which was invested by Hetang Venture Capital, Guofang Innovation and Shanghai Zheyu, and the old shareholder Yuanhe Origin was oversubscribed. This round of financing is mainly used for polishing and landing AI medical agents, and building an academic communication bridge connecting pharmaceutical companies and doctors. In 2022, Enran Venture Capital participated in the Pre-A round of financing of Null Hypothesis Technology as the lead investor.
There are many landing scenes of AI medical care, from medical imaging AI, intelligent auxiliary diagnosis, to AI+ surgical robots, AI pharmaceuticals, etc., which have spawned many popular tracks in the past. However, American unicorn company OpenEvidence and China innovative enterprise "Null Hypothesis" aim at a scene with great social value and closer to commercialization: evidence.
In the clinical process, doctors are usually familiar with common diseases and frequently-occurring diseases, but in the face of intractable diseases, rare diseases or cutting-edge treatments, they often need to quickly and accurately explore evidence from the vast amount of medical literature, guidelines and clinical trials around the world to support clinical decision-making. Therefore, developing high-quality AI productivity tools for it can greatly improve doctors' clinical and scientific research efficiency.
On October 20th, OpenEvidence completed the $200 million Series C financing, which was led by Google Ventures. This year, this startup company, which is positioned as "doctor AI co-pilot", quickly completed three rounds of financing, and its valuation soared from $1 billion at the beginning of the year to $6 billion. Technology giants and star venture capitalists continue to raise money, showing that overseas capital recognizes the potential of the track.
At this stage, many domestic enterprises are also trying to build a doctor AI productivity tool. Based on its years of accumulation in the medical field and in-depth understanding of pharmaceutical companies and doctors, Null Hypothesis, established in 2019, has embarked on a road leading to the same goal as OpenEvidence.
Gu Fei, founder of Null Hypothesis, said, "Many domestic investors regard us as China's' open evidence', and the goal of Null Hypothesis is similar to it, but China's medical ecology is different from that of the United States, and there are obvious differences in doctors' practice habits and drug supervision environment. Therefore, in terms of implementation path, product design, and closed-loop business building, we need to start from the beginning and fully consider the needs of C-end doctors and B-end pharmaceutical companies."
Compared with the United States, the working hours and busyness of doctors in the third-class hospitals in China are generally higher, and the requirements for the ease of use and intelligence of tools are higher. "Doctors need an agent with simple interaction, which can best guess the needs and deliver accurate and comprehensive information, rather than a very complicated retrieval system". In addition, many doctors in China have to take care of clinical and scientific research work at the same time, and the role division is not as clear as that of overseas doctors. "Therefore, it is best for the AI tools provided to doctors to seamlessly connect these two scenarios."
Gu Fei believes that if you directly do the C-end from the beginning, whether it is data accumulation or product polishing, it may be difficult to meet the requirements of head experts. Therefore, in the null hypothesis, the B-terminal feeds back the C-terminal and the C-terminal attracts the B-terminal, that is, first cut into the head pharmaceutical companies with high requirements for content professionalism and compliance, and accumulate high-quality labeling data and iterative algorithms in the process of providing AI tools and services for pharmaceutical companies, and then push 2C's "KnowS" model and agents to the doctor's side.
For AI medical tools, "illusion" is a fatal weakness. Null Hypothesis improves the scientificity of cited evidence through a series of methods. According to Gu Fei, "at this stage, Null Hypothesis has controlled the illusion rate at the level of' expert availability'."
In terms of data collection and management, five years ago, the Null Hypothesis was to set up a part-time doctor team of 100 people to mark the data in detail. Deeper and finer granularity data labeling is the first line of defense to reduce illusion.
At the same time, it has built its own vertical medical database and search engine. When doctors use Null Hypothesis, they do not directly start online search. Instead, medical evidence is obtained in a self-built and daily updated vertical medical database. The database includes literature, guidelines, global medical conference data, clinical trials and even high-quality academic self-media content, so as to give consideration to the accuracy, completeness and novelty of data sources.
In addition, the null hypothesis constructs the AI audit model, which will check the facts of the content before finally providing the doctor with the answer, so as to further reduce the illusion.
After years of deep cultivation in the field of AI medical care, Gu Fei believes that entrepreneurs need to consider a complete business closed loop in the process of product development. For AI medical productivity tools, doctors are high-frequency users, while pharmaceutical companies are important participants and payers. Therefore, understanding the marketing logic and core pain points of pharmaceutical companies is the key.
Through years of cooperation with pharmaceutical companies, Null Hypothesis has clarified a variety of business paths: providing "digital infrastructure" for pharmaceutical companies, that is, using AI to improve the content production and training efficiency of internal medical, marketing and sales teams of pharmaceutical companies; Provide digital marketing and promotion services for pharmaceutical companies. Gu Fei said, "At this stage, the Null Hypothesis has entered the supplier system of more than 30 domestic and foreign head pharmaceutical companies, and most of them are preferred supplier qualifications."
According to reports, with the development of the domestic business model, Null Hypothesis has set its sights on overseas markets such as Europe and Japan, and started its journey to the sea. Its C-end products will also be deeply localized according to the diagnosis and treatment guidelines and doctor habits of various countries.
According to Jost Sullivan's prediction, the scale of China pharmaceutical digital marketing market is expected to increase to 356.8 billion yuan in 2030, and many pharmaceutical companies around the world are increasing their digital marketing expenditure. The market needs professional AI medical tools that can be recognized by doctors and meet the compliance needs of pharmaceutical companies to further improve the efficiency of doctors' diagnosis and treatment and scientific research.
