Skip to content
Mignuti Chatbot

Use case · gp practice

Chatbot for general-practice clinics — opening hours, vaccine info, prescription refills — never diagnoses

The problem

During cold-season, MFAs (medizinische Fachangestellte) field hundreds of phone calls a day — the vast majority are routine: 'when can I get a flu shot', 'is the practice open Saturday', 'I need my Diabetes-Rx renewed'. Phone tree is overloaded, sick patients get busy-signals.

How Mignuti Chatbot solves it

Mignuti Chatbot handles the routine triage: opening hours, walk-in slots, vaccine availability, prescription-renewal request collection. ALL clinical questions ('I have a fever, what should I do?') are hard-blocked and route to 'please call 116 117 (Ärztlicher Bereitschaftsdienst) or come in during walk-in hours'. Designed with the requirements of Heilmittelwerbegesetz + Heilberufe-/Kammerrecht in mind — final compliance check stays with your practice and your Kammer.

Example questions

FAQ

Can the bot diagnose symptoms?

Absolutely not — and this is enforced at three layers: (1) System-prompt blocks any diagnostic intent. (2) Content-policy keywords (pain, fever, symptom, blood, dizzy) trigger immediate handover-message. (3) The fallback message is 'please call 116 117 or come in walk-in hours' — never an attempted diagnosis.

Is this DSGVO-compliant for health-data?

Chat content can contain health data (Art. 9 GDPR) — that demands extra care: PII redaction strips names and identifiers before any AI call or log entry, platform data stays EU-only (Frankfurt), and AI responses run via Anthropic (USA, EU-US DPF + EU SCCs). A DPA (Art. 28) is included; whether Art. 9 processing is permissible for your practice must be assessed with your data protection officer — we provide the documentation for that review.

Can patients book appointments through the bot?

v1.0: bot collects name + insurance + symptom-category + preferred-slot, forwards to your front desk via Slack/email. v2.0 roadmap: direct doctolib + jameda integration.