Shengyi Health and DingTalk jointly advance the implementation of AI models in healthcare, with AI assistant covering over 300 hospitals within 10 days

On December 23, at the AI DingTalk 1.1 new product launch event, Wang Qiangyu, founder of Eseen Health, stated that "Doctor Doukou Super Assistant," developed in collaboration with DingTalk using a large medical industry model, has been adopted by over 300 healthcare institutions across China just ten days after its release on the DingTalk app marketplace. These include major hospitals such as Peking University Third Hospital and Shanghai Hongfangzi Hospital, as well as grassroots facilities like Alashan Left Banner Maternal and Child Health Hospital, demonstrating breakthrough progress in improving diagnostic efficiency and accessibility through medical AI.

Overcoming the "Hallucination" Challenge for Evidence-Based Diagnosis

Wang Qiangyu noted that Eseen Health, a technology company focused on women's health AI, began a joint project with DingTalk six months ago to build a gynecology-specific large model. While the previously launched "Doukou Gynecology Large Model" had already proven technically feasible, this upgraded version—"Doctor Doukou Super Assistant"—further addresses a core pain point in medical AI: traditional general-purpose large models (such as DeepSeek), which rely on RAG technology, still suffer from high hallucination rates that could lead to misdiagnosis risks. Through enhanced domain-specific model training, the Doukou model has achieved a "zero hallucination" goal, ensuring every diagnostic recommendation is grounded in authoritative literature and clinical evidence chains, significantly reducing the risk of misdiagnosis.

Agent + Specialized Models Driving Clinical Adoption of AI

"The integration of models and Agent applications enables professional medical knowledge to truly reach frontline physicians," emphasized Wang Qiangyu. Doctor Doukou Super Assistant not only provides diagnostic suggestions but also automatically generates checklists or references relevant academic literature. The deep integration between its Agent capabilities and specialized models represents a key advancement in empowering healthcare with AI. The DEAP platform provided by DingTalk played a central role in technical implementation. From data annotation and cleaning to algorithm optimization, the DingTalk team worked closely with Eseen Health to overcome technical challenges such as reinforcement learning and reward function design, delivering full-chain support from pre-training to productization.

Industry-Specific Models: The Key to the Second Half of Medical AI

Wang further analyzed industry trends, pointing out that the "second half" of medical AI lies in combining industry-specific models with Agent technologies. General-purpose large models struggle to meet the precision and evidence-based requirements of medical scenarios, whereas industry-specific models—through dedicated training and tool integration—can more efficiently solve real-world clinical problems. DingTalk has already applied similar logic in areas such as AI customer service, validating the broad applicability of industry-specific models. He called on the healthcare sector to embrace this trend: "DingTalk is transforming complex technologies into easy-to-use tools, enabling even small and medium enterprises to access AI capabilities at low cost—this is a critical step toward medical inclusivity."

Building a Trusted AI Ecosystem for Healthcare Inclusivity

Looking ahead, Eseen Health plans to deepen its partnership with DingTalk to expand the application of the Doukou model in gynecological care and health management. Wang Qiangyu said: "The ultimate goal of medical AI is 'evidence-based' and 'trustworthy' decision-making, and DingTalk's AI ecosystem strategy is providing a solid technological foundation for achieving this goal."

We dedicated to serving clients with professional DingTalk solutions. If you'd like to learn more about DingTalk platform applications, feel free to contact our online customer service or email at This email address is being protected from spambots. You need JavaScript enabled to view it.. With a skilled development and operations team and extensive market experience, we’re ready to deliver expert DingTalk services and solutions tailored to your needs!