Deep Intelligent Pharma: How AI Is Reshaping Drug Development Amid Four Funding Rounds in a Year

When a company sets out to build a platform, its leader must be a seasoned veteran who has seen the industry’s ups and downs and can spot opportunities in the cracks. Who would have thought that a pharmaceutical platform could emerge in the new “AI + drug discovery” track? Deep Intelligent Pharma is turning this vision into reality.

On October 9 this year, Deep Intelligent Pharma secured nearly $15 million in Series B funding from Sequoia Capital. This round came just over a year after the company’s founding, during which time Deep Intelligent Pharma had already completed three rounds of financing, attracting high-quality investors such as ZhenFund. What factors have driven such rapid growth for Deep Intelligent Pharma? Let us begin by examining the pharmaceutical industry.

As is well known, drug development is a long and arduous journey. Both century-old multinational pharmaceutical giants and emerging biotech startups frequently incur substantial losses due to R&D failures. Given the risk-averse and profit-seeking business environment in China, it is hardly surprising that 90% of Chinese pharmaceutical companies are generic drug manufacturers.

For Li Xing, founder and CEO of Deep Intelligent Pharma, who spent many years working at multinational pharmaceutical companies and participating in drug development, these bitter fruits seemed endlessly distant. It was not until the emergence of AI that Li Xing appeared to find a way to improve the success rate of drug R&D, accelerate the drug development process, and reduce R&D costs.

Throughout the entire new drug development process, which comprises nearly 20 modules, every stage—from target discovery in the early phases to patient trials during clinical development—is a endeavor heavily reliant on time and capital. The entire timeline ranges from as little as 5–6 years to over a decade. Such prolonged investment cycles further exacerbate risks and costs, while the increased difficulty in internal coordination and review across different stages means that few companies can participate in the full spectrum of drug development. These challenges have given rise to innovative outsourcing models such as CROs (Contract Research Organizations) and CMOs (Contract Manufacturing Organizations), where specific modules of the R&D process are delegated to specialized firms to mitigate risks and accelerate capital recovery.

Furthermore, the diversity of diseases and the specificity of drugs have created a complex demand market. For common diseases, sufficient data is available, resulting in lower R&D difficulty and costs, as well as high product demand. In contrast, while the overall market for rare diseases is large, the scarcity of clinical cases for any single rare disease makes R&D challenging. Moreover, the market size for each specific product post-launch is limited, deterring many enterprises from entering this field.

Meanwhile, the mutable nature of viral characteristics and drug crystal forms can lead to delays, loss of efficacy, or failure even after drug development is completed. For instance, Abbott’s Ritonavir experienced a change in its crystal form after market launch, resulting in a $251 million loss. Pfizer invested heavily in research on Parkinson’s disease and Alzheimer’s disease and, together with several other pharmaceutical companies such as GlaxoSmithKline and Eli Lilly, collaborated with government agencies to establish the Alzheimer’s Disease Research Fund in 2015. However, on January 7, 2018, Pfizer announced it would gradually discontinue the development of new drugs for treating Alzheimer’s and Parkinson’s diseases.

Finally, as traditional drug development relies on continuous trial and error, when companies aim to accelerate R&D progress and enter the market as quickly as possible, a surge in labor costs is inevitable.

Following the emergence of AI technology, Li Xing developed a strong interest in it and began studying AI in his spare time. In March 2017, he outlined the concept of applying AI to drug discovery, marking a turning point that propelled him to delve deeply into exploring how AI technology could be utilized in pharmaceutical development.

Li Xing told a VCBeat reporter that the core of “AI + New Drug Development” lies in leveraging artificial intelligence algorithms to significantly reduce the costs associated with target discovery and molecular trial-and-error, while utilizing advantages such as digitalization to shorten the entire new drug R&D process. According to VCBeat’s compilation, this can be summarized into the following three directions.

Researchers leverage AI to analyze hundreds of millions of standardized and annotated biological, pharmacological, and clinical data points. This enables the identification of candidate drugs and biomarkers, as well as accurate early-stage predictions of crystal forms that influence drug properties. Such capabilities assist pharmaceutical companies in selecting the most promising molecular candidates, crystal forms, solid phases, and R&D pathways at an early stage, thereby choosing the lowest-risk routes and accelerating development. AI enterprises can also reduce the number of required trials by predicting toxic side effects. Furthermore, AI effectively extracts valuable insights from massive volumes of gene data characterized by single structures.

AI can significantly improve the research and development of raw material synthesis by analyzing vast amounts of data and literature. Optimized processes can reduce the number of subsequent experiments and, in some cases, lower raw material costs. After obtaining pharmacokinetic data, AI can more precisely guide clinical studies, including determining dosing frequency and dosage during clinical administration.

By leveraging a clinical data integration system, enterprises assist hospitals and departments in establishing structured electronic medical record (EMR) databases, thereby enhancing efficiency across diagnosis, follow-up, and scientific research. With such extensive information interconnected, AI can help physicians identify patients eligible for clinical trials, thus accelerating new drug development starting from the clinical trial phase.

In light of the high R&D costs and other challenges associated with the complex new drug development process that domestic pharmaceutical companies struggle to bear, Deep Intelligent Pharma leverages advanced artificial intelligence technology to empower the entire workflow of new drug development and regulatory submission, helping pharmaceutical enterprises in China and worldwide develop new drugs faster and more effectively.

Deep Intelligent Pharma adheres to a customer-centric approach, leveraging AI to accelerate target discovery and clinical research, thereby building foundational platforms for pharmaceutical companies. The company has dedicated substantial human resources in equal measure to both of these areas.

Building a foundational platform is no overnight endeavor; establishing its knowledge graph alone requires substantial time. To this end, Deep Intelligent Pharma has recruited numerous translators to translate overseas literature, aiming to develop products better aligned with user needs for the domestic market.

Deep Intelligent Pharma has successively launched multiple product prototypes, including AI-driven drug synthesis, drug design, and drug activity prediction. For the clinical research phase, it has introduced an AI-driven pharmacovigilance system, a regulatory affairs system, and a translation platform, among others. These achievements cover key nodes across the entire new drug R&D process. Notably, its pharmacovigilance system has received widespread acclaim from users.

In the area of target discovery, Deep Intelligent Pharma is currently primarily focused on oncology and rare diseases. Li Xing revealed, “We can tailor our algorithms to meet clients’ specific needs.”

Deep Intelligent Pharma’s next-generation AI-driven drug synthesis system, “Bodhi,” significantly enhances chemists’ productivity by extensively learning from public patent and literature databases. Upon inputting a chemical structure, the system instantly recommends multiple optimized synthetic routes. The computational power supporting this algorithm is derived partly from self-built infrastructure and partly from public cloud services. This hybrid approach enables Deep Intelligent Pharma to handle concurrent multi-task workflows and allocate computational resources according to task priority.

However, despite possessing advanced compound screening algorithms, Deep Intelligent Pharma rarely engages in proactive drug discovery. Li Xing stated, “For Deep Intelligent Pharma, leveraging algorithms to discover drugs can effectively enhance efficiency and success rates, but this is not a sustainable model. While we might achieve exceptional revenues in the short term, in the long run, our greatest significance lies in focusing on our core competencies—enabling more pharmaceutical R&D institutions and scientists to utilize our technology, taking Chinese technology global, and ultimately benefiting more patients.”

As data volumes grow, the efficiency of the Deep Intelligent Pharma foundational platform will continue to improve. For pharmaceutical companies, relying solely on manual processes for target screening, clinical trials, and regulatory approval for market launch is time-consuming, labor-intensive, and costly. In this context, Deep Intelligent Pharma offers an optimal solution.

Many people long for a spontaneous trip, but not everyone has the courage to embark on one. For Deep Intelligent Pharma, this past year has been an arduous journey defined by letting go and rebirth.

Li Star granted the interview during a lunch break between meetings. The past few weeks have been a critical period for Deep Intelligent Pharma, with its Series B financing just concluded and several clients extending olive branches. As CEO, Li Star must not only address client demands but also ensure the smooth operation of the company itself. Yet time is carved out; those who recognize scarcity are adept at uncovering scarce resources.

Li Xing earned her Master’s degree in Pharmaceutics from Peking University in 2006 and immediately embarked on a 12-year career in drug development at multinational pharmaceutical companies such as Pfizer, Sanofi, and Johnson & Johnson. With the widespread attention garnered by events like AlphaGo’s victory, Li became deeply fascinated with artificial intelligence and began self-studying AI technologies. In March 2017, she recognized that AI represented an entirely new track with the potential to disrupt drug research and development. By October of that year, she decisively left behind the lucrative compensation and comfortable environment of multinational pharmaceutical firms to found Deep Intelligent Pharma Co., Ltd. in Beijing.

Talent constitutes the robust tendons of an enterprise, while culture serves as its immortal soul. The development of Deep Intelligent Pharma to date has been inseparable from the support of talent and culture. The drug R&D process involves approximately 20 functional modules; even experts in a specific module typically possess only a general understanding of other modules. Furthermore, the pharmaceutical industry remains distinct from the IT and internet sectors. This disparity presents the greatest current challenge in the AI-driven drug discovery sector: the effective integration of pharmaceutical experts and AI engineers, or alternatively, the acquisition of cross-disciplinary talent combining industry expertise with technical proficiency.

Deep Intelligent Pharma has recruited a large cohort of talent from China’s leading internet companies, while also maintaining a team of professionals from the new drug R&D departments of multinational pharmaceutical firms. The integration of these two distinct talent pools has become Li Xing’s most pressing challenge at present. Top-tier talents often possess their own distinctive ethos: pharmaceutical professionals exhibit the caution characteristic of the industry, whereas internet professionals display a fearless, aggressive drive.

In response, Deep Intelligent Pharma has demonstrated a clear business model aimed at maximizing customer satisfaction and ensuring successful product implementation. Guided by this model, although employees at Deep Intelligent Pharma maintain independent thinking, they have aligned around shared goals and exerted collective efforts. Li Xing stated, “Deep Intelligent Pharma’s talent structure is now highly comprehensive. After long-term collaboration and integration, some pharmaceutical experts have successfully transitioned into roles such as AI product managers, data mining engineers, or AI coaches.”

In July of this year, Deep Intelligent Pharma had only slightly more than 20 clients. Within less than six months, that number rose to over 50. This growth is inseparable from Deep Intelligent Pharma’s aggressive, results-driven corporate culture. Li Xing often works alongside employees until the early hours of the morning, while also actively participating in various seminars to promote Deep Intelligent Pharma’s products as he continues to learn. Meanwhile, product quality serves as the foundation for its sales success. Li Xing stated, “The market environment in the field of new drug development is relatively homogeneous; as long as our products can help enterprises improve efficiency, they will adopt them.”

Currently, Deep Intelligent Pharma’s AI pharmacovigilance system and AI-powered automated translation submission system have achieved large-scale commercialization; its AI-driven early drug discovery platform and AI-assisted organic synthesis system (for chemical synthesis and process improvement) are on the verge of commercialization.

In the future, Deep Intelligent Pharma will continue to deepen its engagement in drug R&D, striving to launch more groundbreaking “AI + New Drug Development” products through independent research and development, so as to more effectively enhance the efficiency and success rate of drug discovery and more efficiently drive the development of the pharmaceutical industry.