AI Voice Chatbot with ElevenLabs & OpenAI for Restaurants

This n8n workflow creates an interactive voice-based chatbot using ElevenLabs for voice interactions, OpenAI for intelligent responses, and Qdrant vector database for RAG (Retrieval-Augmented Generation). It starts with a webhook from ElevenLabs capturing user voice queries, processes them via an AI Agent with tools for vector store retrieval, generates context-aware text responses, and sends them back to ElevenLabs for TTS delivery. Window Buffer Memory maintains conversation history for natural, ongoing dialogues.

Key benefits include hands-free customer support, personalized service without human agents, and scalability for high-volume interactions. It saves time by automating responses from a knowledge base, reducing errors, and enabling 24/7 availability. Use cases span restaurants (e.g., menu queries, reservations), retail shops (product info, orders), and customer service hotlines.

Setup involves creating an ElevenLabs agent, configuring n8n credentials for OpenAI, Qdrant, and webhooks, defining system prompts (e.g., as a waiter), and deploying the workflow. It's highly customizable for specific domains, enhancing engagement with realistic voice synthesis and contextual memory.