PSKin: How a Beauty Brand Automated LINE Support with AI

The Challenge

PSKin (彭賢禮皮膚科診所) is a dermatology and medical aesthetics clinic in Kaohsiung, Taiwan, operating two locations with 9 specialist dermatologists. Their services range from general dermatology to laser treatments, hair transplants, and cosmetic procedures.

In Taiwan, LINE reaches 95% of smartphone users. PSKin’s LINE Official Account had become the primary channel for patient communication — appointment bookings, clinic hours inquiries, treatment questions, post-procedure care, and doctor recommendations.

The problem was volume and repetition. Front desk staff fielded hundreds of LINE messages daily, and the vast majority were the same questions:

• “What are your clinic hours?”
• “How do I book an appointment?”
• “Which doctor should I see for acne?”
• “What’s the registration fee?”
• “How do I care for my skin after mole removal?”

Staff spent hours on repetitive responses — time that should have gone to in-person patient care. After-hours messages went unanswered until the next business day, leading to missed appointments and frustrated patients.

The Solution

PSKin deployed JieGou’s Chat Agent connected to their LINE Official Account, powered by a comprehensive knowledge base built from the clinic’s own FAQ documentation.

Knowledge-Base Grounding, Not AI Hallucination

The foundation is a curated knowledge base covering clinic information, appointment flows, fee structures, operating hours, all 9 doctors with specializations, post-treatment care protocols, and administrative policies.

Every AI response is grounded in this approved knowledge base — not generated from the model’s training data. In a medical context, this is critical: patients get accurate clinic-specific information, not generic health advice.

How It Works

Patient sends LINE message
  → JieGou LINE MCP Server receives webhook
  → Chat Agent processes with knowledge base context
  → AI generates grounded response
  → Response sent via LINE
  → Low confidence? → Escalate to staff with context

The chat agent handles clinic logistics, appointment booking and modification, doctor recommendations, post-procedure care instructions, and payment questions. Questions outside the knowledge base are escalated to staff with full context.

The Results

Operational Impact

• 24/7 coverage — Patients get instant responses at 3 AM, on Sundays, and during holidays
• FAQ deflection — Common questions handled by AI without staff involvement
• Staff time recovered — Front desk spends less time on repetitive LINE responses, more on in-person care
• Zero additional headcount — Round-the-clock support without hiring

Patient Experience

• Instant responses — Answers in seconds instead of hours
• Accurate information — Grounded in clinic-approved data: correct phone numbers, accurate hours, proper doctor recommendations
• Natural language — Conversational Traditional Chinese responses matching the clinic’s friendly brand voice
• Seamless escalation — Complex questions routed to staff with context, so patients don’t repeat themselves

Governance & Trust

• Knowledge-base grounding — No hallucinated medical advice
• Audit trail — Every interaction logged for quality review
• Human escalation — AI routes edge cases to staff rather than guessing

Why JieGou

PSKin chose JieGou for five reasons:

1. LINE-native integration — Direct webhook connection to LINE Official Accounts via JieGou’s LINE MCP server, no middleware required
2. Knowledge base control — Full control over what the AI can reference, with easy updates when hours or procedures change
3. Traditional Chinese support — Full zh-TW support across the platform, knowledge base, and AI responses
4. Department-first setup — Pre-configured support pack connected to their knowledge base, setup in hours not weeks
5. Built-in governance — Audit logging, escalation rules, and knowledge-base grounding are structural, not afterthoughts — critical for a medical practice

The Pattern

PSKin’s case demonstrates something applicable beyond healthcare: messaging-channel customer support is a solved problem when you combine channel-native AI with curated knowledge bases and governance controls.

The ingredients: a knowledge base reflecting your business, an AI agent constrained to it, escalation paths for edge cases, audit trails for compliance, and native integration with the channel your customers use.

For PSKin, that channel is LINE. For other businesses, it might be WhatsApp, Instagram DMs, or Facebook Messenger. The architecture is the same.

---

PSKin (彭賢禮皮膚科診所) is a dermatology and medical aesthetics clinic in Kaohsiung, Taiwan. They use JieGou for LINE customer support automation.