【China】So-Young Founder Jin Xing: AI Deployment in Chain Clinics Must First Clear the “Data Governance” Hurdle

At the 2026 World Artificial Intelligence Conference, So-Young founder Jin Xing revealed that the company is advancing AI transformation across its 50-plus light medical aesthetics clinics in China. However, the core work is not direct AI application but rather a prolonged phase of “data governance.” Since 2023, So-Young has deployed over 100 engineers to capture data through voice-to-text conversion, integration with diagnostic device interfaces, ERP system tracking, and video recording. This effort has accumulated over 3 million comparison photos and 530,000 follow-up reports across 1.75 million treatments. Yet ensuring data authenticity has emerged as the primary challenge. For instance, when tracking customer wait times, methods like tablet check-ins, wristbands, Bluetooth, and Wi-Fi all encountered issues ranging from altered employee behavior to disrupted customer experience, ultimately requiring a multi-technology combination to reconstruct real movement patterns. For treatment inspections, So-Young plans to launch an AI intelligent inspection system by late July, using visual recognition for 24/7 SOP compliance monitoring. However, this first requires upgrading network bandwidth to 200 Mbps at every clinic nationwide. Jin believes AI will not immediately make individual clinics more profitable in the short term, but it can expand the management radius of chain organizations, transforming personal experience into replicable systemic capabilities—all predicated on first converting the real world into machine-readable data.

At the recently concluded 2026 World Artificial Intelligence Conference, enterprise AI emerged as one of the most closely watched sectors. Unlike previous years that focused on showcasing model parameters and generative capabilities, this year’s exhibition featured more AI agents, industry solutions, and enterprise deployment tools, as AI companies attempt to extend their reach into the deeper waters connecting enterprise data, systems, and business processes. However, when AI truly steps into the physical world—especially in scenarios like chain clinics that heavily rely on offline interactions—things are far less smooth than product demonstrations suggest.

Which consultation room did the customer enter? What procedures did the doctor perform? Where was the equipment moved? Why was a medication dispensed and then returned? These things do not automatically become data that machines can understand. In a recent in-depth, nearly two-hour exchange with media including

So-Young is advancing AI transformation across its 50-plus light medical aesthetics clinics in China, aiming to apply AI to consultations, treatment quality inspections, customer triage, and clinic operations. But what left a lasting impression was not the AI applications themselves, but rather how much foundational work—seemingly unrelated to AI—the company had to do first just to get AI into the clinics.

For chain enterprises, one of AI’s most enticing values lies in replicating the capabilities of top-performing clinics and employees. A single clinic can sustain operations relying on its manager, doctors, or skilled staff, but when the number of clinics grows from 10 to dozens or even hundreds, the enterprise must ensure all locations operate to the same standards. So-Young aims to build a standardized light medical aesthetics chain. According to Jin’s vision, the same customer with the same needs, visiting different clinics and consulting different doctors and advisors, should receive highly similar treatment plans. The reality, however, is that the treatment plan a customer ultimately receives still largely depends on the individual knowledge and experience of the doctor and consultant.

For AI to narrow this gap, it first needs to know what actually happens inside the clinics. A complete medical aesthetics treatment includes at minimum: the customer stating their needs, undergoing skin diagnostics, the doctor identifying issues, designing a treatment plan, preparing medications and consumables, completing the treatment, and post-treatment follow-up testing and feedback. However, this information is scattered across various media: the customer’s needs are captured in a consultation conversation, skin condition data comes from diagnostic devices, medication and dosage records reside in the supply chain system, the treatment process is a video recording, and the final outcome must be assessed through before-and-after photos, follow-up reports, and reviews.

So-Young adopted different collection methods for different data types. Consultation conversations are recorded via voice-to-text; skin diagnostic devices are integrated with backend interfaces to break down previously view-only full reports into structured indicators; medications and consumables are tracked through the ERP system; and the doctor’s treatment process is captured on video. According to Jin, So-Young has accumulated over 3 million before-and-after treatment comparison photos and nearly 530,000 follow-up reports across approximately 1.75 million treatments.