Build a Chatbot, Voice Agent and Phone Agent with Voiceflow, Google Calendar and RAG

This n8n workflow automates the creation of a comprehensive voice and chat assistant integrated with Google Calendar, a Qdrant vector store, and a retrieval-augmented generation (RAG) system. It enables interaction via Voiceflow, supporting voice, chat, and phone agents for customer support, appointment scheduling, and knowledge retrieval.

The workflow begins with three main webhooks for input collection: ‘n8n_order’ for order tracking, ‘n8n_appointment’ for scheduling, and ‘n8n_rag’ for knowledge queries. Incoming data from these webhooks trigger the workflow. The order and appointment nodes are used for tracking order status and scheduling appointments, updating Google Calendar accordingly by creating events with dynamic summaries and descriptions.

Simultaneously, the workflow manages a Qdrant vector store, creating collections, deleting points for refreshes, and integrating with Google Drive to download and process documents for embedding. These embeddings are stored in Qdrant for efficient retrieval.

For the knowledge retrieval component, the workflow employs a RAG system, retrieving relevant information about a company or product from the vector store, and using OpenAI’s GPT-4 mini models for conversational responses. The responses are then sent back via webhook responses, providing real-time feedback or answer generation.

Overall, this workflow is ideal for businesses seeking to provide automated, voice-activated customer service, appointment management, and intelligent knowledge retrieval—enhancing engagement and operational efficiency.