The AI operating system for Korean medicine clinics.
HaniOps runs the back office and the growth engine of a clinic: patient follow-up, herbal prescriptions, staff operations, and an AI pipeline that keeps the clinic visible on search engines and AI assistants. It was built inside a working clinic by the physician who runs it, and it is used there every day.
Small clinics carry enterprise-sized workloads.
Korea has about 14,000 Korean medicine clinics. Most are run by one or two physicians with a handful of staff, and the physician is also the marketer, the operations manager and the IT department.
- Follow-up falls through the cracks. Herbal medicine runs in multi-week courses. Pickup calls, check-in calls and refill reminders are tracked in notebooks or spreadsheets, and missed calls mean lost patients.
- Generic EMR tools stop at the chart. They do not handle decoction schedules, herbal inventory, staff leave or the clinic's own workflow rules.
- Discovery is moving to AI. Patients now ask ChatGPT, Gemini and Google's AI Overviews which clinic to visit. A clinic that publishes nothing structured simply does not appear in those answers.
What we build
Two products in production and one in development, sharing the same clinic data.
HaniOps Clinic
Multi-tenant clinic management SaaS.
- Rule-based call worklists for pickup, check-in and refill
- Herbal prescriptions and decoction calendar
- Messaging scenarios (SMS / KakaoTalk)
- Staff leave, tasks, inventory and purchase orders
- Reservation sync with the clinic's existing EMR
- Owner dashboard: new patients, revenue, retention
HaniOps Growth
An agent-driven website and content engine.
- AI agents draft daily articles; the physician approves before publishing
- Medical-advertising compliance checks on every draft
- Structured data so search engines and AI assistants can cite the clinic
- Weekly measurement of the clinic's share of voice in ChatGPT, Gemini, Perplexity and Google AI Overviews
Haru Hanyak
A patient-facing app for people on a herbal medicine course.
- Daily dosing and progress check-ins
- Weight-management program support
- Data flows back to the clinic's HaniOps workspace
Inside the product
Screens from HaniOps Clinic. The interface is in Korean for Korean clinics; these captures use a demo clinic with fictional patients.
Traction
Everything below runs in production today. Our first customer is our own clinic, which is why the product is shaped by real daily use rather than assumptions.
Figures as of September–October 2026. Patient and prescription counts are aggregate record counts; no patient data leaves the clinic workspace.
Built with AI, for AI-era clinics
We are a very small team, so AI is not a feature on top of the product. It is how the product gets built and how it delivers value.
AI-native engineering
The founder ships the backend (Go, PostgreSQL), web apps (Next.js, Astro) and infrastructure (Cloud Run, Vercel, Cloudflare) by working with coding agents, which is how a practicing physician sustains a weekly release cadence.
Agents in the workflow
Agents plan topics, draft articles, generate images, check medical-advertising rules and prepare structured data. A human physician remains the final approver for anything a patient will read.
What we would do with more AI capacity
Roll the content and visibility engine out to every clinic on the platform, add an in-app assistant that answers staff questions from clinic data, and give patients in Haru Hanyak a safe, physician-supervised companion.
Team
Bong-kyu Park, KMD, PhD — Founder
Licensed Korean medicine doctor and director of Kyung Hee Meerae Korean Medicine Clinic (Nampo, Busan). PhD in Korean Medicine, Kyung Hee University (2026). Runs the clinic and builds HaniOps himself, so every feature is tested on real patients' workflows before it reaches another clinic.
Company
- Company
- Yuonjae
- Founded
- 2026
- Products
- HaniOps · Haru Hanyak
- Location
- Busan, Republic of Korea
- Contact
- [email protected]