For anyone evaluating us
In 2018, we took Ta-Tung Hospital's health checkup bookings — kept in a paper book worn to shreds — online, and 80% of reservations stopped going through the phone. The center's then-director recommended us to the next hospital. Then the next. Eight years later, the health records one person leaves behind — at the checkup center, at work, in the community — flow into one place for the first time. The story is only halfway through.
Figures cited from the hospital interviews in Watch Magazine No. 276 (May 2026).
Why this compounds
Every institution we serve teaches us one more kind of frontline. Every frontline we understand connects one more segment of a person's health record. The more complete the record, the better we know who needs what help, and when. The more precise the help, the more visible the outcome — and the outcome is what makes the next institution say yes.
The first half of this loop is proven by eight years in hospitals; the second half is happening now in enterprise settings.
Every turn,faster than the last
Why the flywheel keeps turning
The ecosystem doesn't scale on headcount — it scales because these three engines cut the work each loop needs (structuring, judging, filing) to the minimum. They run on one shared foundation, and every record feeds all three at once.
one-pass import
Turns spreadsheets, scanned reports, and forms in every format into one structured, traceable medical record (FHIR).
Ends “I said it, now I have to type it again — and the next batch is already here.”
population insight
Lifts individual records to a group view, surfacing the cohorts worth acting on — which high-risk group needs follow-up now — and turns them straight into care plans. Employers see only the group structure, never individuals.
Ends “the data arrived, but no one can see the whole picture.”
administrative paperwork
OCR plus document understanding to auto-generate the government-format forms, audit records, and compliance documents.
Ends “we only found the gap right before the audit” — compliance paperwork goes from days to minutes.
Others would rebuild four systems and train four sets of AI. We grew four settings from one foundation, and every record feeds the same three engines — the data gets richer, the models get sharper. That's the time gap others can't close.
Why us, and not a bigger company
Because this loop has no shortcut. Trust on the medical frontline, the different report format of every checkup center, every government-mandated form — they can only be bought with time. Our clock started in 2013 and has never stopped.
AI makes all of it run faster: speech becomes records the moment it's spoken, a photographed report files itself, compliance paperwork shrinks from days to minutes. But AI is only the accelerator. The steering wheel is in the hands of eight years of frontline experience.
80%
of checkup bookings moved online (Ta-Tung Hospital)
100+
premium checkups per month, up from 20–30 (Siaogang Hospital)
2013
the year we entered the medical frontline — and never left
The first two figures are from hospital interviews in Watch Magazine No. 276.
Where this goes
From one health management center, referred into the entire KMU hospital system. Tools for four care settings running in real institutions, with MOHW certification and selection. What this stage proved: the medical frontline is willing to hand us its daily work.
Checkup centers and corporate workplaces are two ends of the same population — and we stand at both. The Healthy Taiwan initiative is opening hospital budgets and willingness right now. Policy windows don't stay open forever; that is why we are accelerating at this moment.
Once records are complete enough, the red flags on a checkup report go find the right service on their own. And what can be matched goes well beyond health services: fitness, stress relief, hands-on workshops, family days, learning, and counseling when it's needed. Employees choose for themselves, employers only see whether people are getting healthier, and providers reach the people who truly need them. A business all three sides pay for gladly is a business that grows on its own.
Our first turn of the loop took five years, because every step was manual. Now AI is compressing every turn shorter and shorter.
How big the market for the third stage is: Taiwan has over 1.67 million enterprises and more than 9 million employees, and most of the employee checkups they run each year end up in a drawer. Turning "in the drawer" into "connected to a service" is the size of this business.
This business is charging and running today: every partner institution is a paying customer — the annual service fee converts into AI usage credits, updates are free, and customers can deploy in their own environment. Later, the matching itself becomes revenue.