At 21:18 on 11 November, a 36-year-old woman identified only as Ms Li sent her annual physical-exam report to an AI doctor named Dawei (大为). The report had several abnormal markers, and she was anxious about them. Within minutes the assistant had walked through each marker, suggested next steps for follow-up, and assembled the follow-ups, exercise adjustments and nutrition notes into a checklist she could act on, according to JD Health's campaign report.
That vignette, relayed in JD Health's own campaign report, is the consumer-facing headline. The rest of the story is where AI landed during China's biggest shopping festival — and some of it happened far from any shopping cart.
Key takeaways
- Consultation orders for JD Health's AI doctor Dawei grew more than 5 times year on year during the 11.11 campaign period (JD Health campaign report carried by Xinhua-affiliated and Jimei media, 11–12 November 2026).
- The flagship use cases were physical-exam report interpretation, health-consultation triage, medication reminders and daily health management — with an AI inquiry volume up 130% year on year in the nutrition channel, which generated over 1 billion yuan (about US$139 million) in guided sales.
- JD's auto AI steward diagnosed dashboard warning lights for car owners online during the same window.
- In eldercare settings, devices that understand Henan dialect served residents of a Xuchang nursing home, an AI cooking robot cooked for more than 100 elderly people at a charity canteen in Henan's Fanying village, and a JD Logistics drone crossed mountains in 7 minutes to deliver parcels to villagers in Sichuan (Jimu News, 11–12 November 2026).
- Campaign-wide, JD reported more than 50 million AI products sold and over 300 million service orders through 23:59 on 11 November — figures the platform self-reports.
The doctor's visit that starts with a PDF
Dawei's growth number — consultation orders up more than fivefold — is the sharpest single metric in JD Health's report, and the exam-report interpretation is the use case doing most of the work. This is a well-chosen wedge. An annual physical is a document most people receive once a year, cannot fully read, and feel compelled to do something about; the demand is recurring, seasonal and emotionally urgent, which is a rare combination for a consumer AI product.
The report describes the workflow: upload the report, get marker-by-marker interpretation and medical advice, receive an action list covering follow-up visits, exercise and nutrition, then continue into online consultation and purchase if needed. Whether that journey ends in care or ends in a shopping cart is the part the campaign report does not separate — which is the point of a campaign report.
JD Health's own numbers frame the scale. Around the same window it reported its nurse-at-home accompaniment service growing 240% year on year, its examination-centre user base up 300%, and an online hospital network of top-tier tertiary doctors handling rapid consultations at 3.3 times last year's volume. The nutrition channel's AI inquiry volume grew 130%, and purchases made under nutritionist guidance exceeded 1 billion yuan (about US$139 million), up more than 260%. None of these distinguish AI-driven demand from ordinary campaign uplift, so the honest read is directional: AI sat in the path of a large volume of health-related demand during the festival.
Warning lights, dialects and dinner
Two other AI applications in JD's report work in less marketable settings.
The first is the car AI steward, which spent 11.11 evening helping owners interpret dashboard warning lights online — triage for the moment a driver notices an unfamiliar symbol and the nearest garage is closed.
The second cluster is eldercare and rural service, reported by Jimu News:
- In a Xuchang nursing home, AI devices built to understand Henan dialect (河南乡音) serve as daily companions — a reminder that for elderly Chinese users, speech interfaces fail first on accent, and dialect-tuned models are the difference between a usable product and a toy.
- At a charity canteen in Fanying village, Henan, an AI cooking robot prepared meals for more than 100 elderly residents, taking over the peak-hour labour that village canteens run on volunteers.
- In mountainous Sichuan, a JD Logistics drone covered a mountain crossing in 7 minutes to deliver parcels to villagers' doorsteps — a route that would take far longer by road.
These are small deployments. Their significance is distributional: AI services showed up in places consumer-tech coverage rarely visits, solving problems — dialect comprehension, canteen labour, mountain logistics — that are specific and unglamorous.
The context numbers, and what they don't say
JD's campaign totals — over 50 million AI products sold, more than 300 million service orders, over 70 million interactions for its JoyInside device platform, and a JoyAI app daily-active peak at five times normal — set the backdrop. A separate detail: first-time participants in the festival averaged 6 orders each, suggesting the AI-guided shopping path (voice-ordering through a digital assistant while holding a baby, in the report's example) is pulling in users who were not shopping this way before.
But every figure here is platform self-reporting. There is no external audit of what counts as an "AI product", how "consultation orders" are defined, or what baseline last year's comparison used. The fivefold growth of Dawei's orders is best read as evidence that AI consultations have moved from novelty to a routine entry point for JD's health funnel — the absolute volume stays undisclosed.
Honest limitations
This article's numbers come from JD's own campaign reports and from Jimu News, a Hubei-state-media outlet relaying platform-provided material. None of the growth figures are audited, baselines are not published, and "AI product" sales include AI-adjacent hardware categories. The anecdotes — Ms Li's exam report, the Xuchang nursing home, the Fanying canteen — are single-site illustrations supplied for campaign purposes, not samples. Dawei is an AI health assistant operating within an online-consultation platform, not a licensed diagnostician; its advice routes users toward clinicians, and the platform's marketing does not claim otherwise. Finally, 11.11 is a promotional peak: growth rates during it overstate steady-state demand by definition.
What readers can do now
- If you build consumer health AI, note the wedge that worked: a recurring, high-anxiety document (the annual physical) rather than open-ended symptom chat, with a structured action list as the output.
- If you design voice interfaces for elderly users, treat dialect coverage as a launch requirement, not a roadmap item — the Xuchang deployment implies Mandarin-only products are non-starters in regional markets.
- If you evaluate e-commerce AI claims, apply the same three questions to any campaign report: what counts as an "AI order", what is the baseline, and what share of the funnel ends in a purchase versus a service.
