RAG-Powered WhatsApp AI Agent for Private Medical Centre
Built a WhatsApp-based AI agent for a private medical centre that answers patient questions using retrieval-augmented generation. The system pulls accurate answers from the clinic's own knowledge base instead of hallucinating responses.
Business Challenge
The clinic's front desk handled repetitive patient inquiries manually, including appointment procedures, service availability, and clinic policies. Staff time was consumed by questions that required no clinical judgment.
Solution Delivered
Deployed an n8n workflow connecting GreenAPI for WhatsApp messaging, Pinecone for vector search, Gemini for embedding and response generation, and Airtable for logging. Patient messages trigger a RAG pipeline that retrieves relevant clinic documents and returns accurate, context-aware responses automatically.
Outcomes Achieved
Eliminated manual handling of routine patient inquiries. Front desk staff time redirected to clinical support tasks. Response time reduced from minutes to under 5 seconds.
Measurable Business Outcome
Reduced front desk response time for routine inquiries from an average of 4 minutes to under 5 seconds. Automated 100% of FAQ-type patient messages through WhatsApp.