An illustrative scenario of how we would work.
The challenge
Imagine a clinic whose reception spends much of the day answering the same questions by phone, WhatsApp and email: opening hours, prices, how to prepare for a test or how to change an appointment. Meanwhile, patients at the desk are kept waiting.
The solution
We would start by gathering the real questions the clinic receives and writing approved answers together. The assistant would answer only from that information, on the website and on WhatsApp.
It wouldn’t give medical advice or look at patient records: faced with a clinical question, an emergency or anything it can’t resolve, it would hand the conversation over to reception with a summary. For appointments, it would connect to the clinic’s existing booking system, if that system allows it.
Before opening it to patients, we would test it with real questions and measure how often it gets them right. The provider and settings would be chosen to give health data the protection it requires.
Technologies
- OpenAI
- Claude (Anthropic)
- n8n
- WhatsApp Business API
- Website chat
The result
Reception should spend less time on repeated questions and more on the patients in front of them. Patients would get answers outside opening hours too, and the clinic would learn what people ask most so it can improve its information.