Build a Voice RAG Chatbot with ElevenLabs and OpenAI
detail.loadingPreview
Create an interactive voice-enabled RAG chatbot using ElevenLabs for speech synthesis and OpenAI for AI agent capabilities. This workflow integrates with Qdrant Vector Store and Google Drive for knowledge retrieval, enabling intelligent responses to user queries.
🚀Ready to Deploy This Workflow?
About This Workflow
Overview
This n8n workflow orchestrates a sophisticated RAG (Retrieval Augmented Generation) chatbot that can understand and respond to voice input. It leverages ElevenLabs for converting text-to-speech and speech-to-text, and OpenAI's language models to power an AI agent. The core of the RAG system involves retrieving information from a knowledge base stored in a Qdrant Vector Store, which is populated with documents downloaded from Google Drive. The workflow is structured into distinct steps: setting up an agent on ElevenLabs, creating and populating the Qdrant collection, vectorizing documents, and finally, enabling the AI agent to process queries and generate voice responses.
Key Features
- Integrates ElevenLabs for voice input/output and OpenAI for AI processing.
- Implements RAG architecture using Qdrant Vector Store for knowledge retrieval.
- Fetches documents from Google Drive for the knowledge base.
- Allows for a customizable AI agent with specific system prompts and tools.
- Provides a step-by-step guide for setup and configuration.
How To Use
- ElevenLabs Setup (Step 1):
- Create an agent on ElevenLabs (e.g.,
test_n8n). - Configure the 'First message' and 'System Prompt' to define the agent's persona and instructions.
- Set up a webhook tool (e.g.,
test_chatbot_elevenlabs) with a description and the n8n webhook URL. - Enable 'Body Parameters' and define a 'question' field for user input.
- Create an agent on ElevenLabs (e.g.,
- Qdrant Collection Setup (Step 2):
- Ensure your Qdrant URL and collection name are correctly configured in the
Create collectionandRefresh collectionHTTP Request nodes. - You may need to adjust the
Create collectionandRefresh collectionnodes based on your Qdrant setup (e.g., usingCOLLECTIONas a variable).
- Ensure your Qdrant URL and collection name are correctly configured in the
- Document Vectorization (Step 3):
- Configure the
Get foldernode to point to your Google Drive folder containing the documents. - The
Download Filesnode will download these documents. - The
Default Data Loader,Token Splitter,Embeddings OpenAI, andQdrant Vector Store(insert mode) nodes will process, embed, and store your documents in Qdrant. - Ensure the
QDRANTURL,COLLECTION, and OpenAI API credentials are set.
- Configure the
- RAG System and Voice Interaction (Step 4):
- The
AI Agentnode (using OpenAI) will receive questions from the ElevenLabs webhook. - The
Vector Store ToolandQdrant Vector Store(retrieval mode) nodes will fetch relevant information from your knowledge base. - The
OpenAInode (LM Chat OpenAI) will generate a response based on the retrieved context. - The
Respond to ElevenLabsnode sends the text response back to ElevenLabs for voice synthesis.
- The
- Testing:
- Click 'Test workflow' in n8n.
- Click 'Test AI agent' in ElevenLabs.
- Ask a question to your voice agent; the response should be synthesized by ElevenLabs.
Apps Used
Workflow JSON
{
"id": "a6d52c63-42a3-4081-b701-247620b5c6c4",
"name": "Build a Voice RAG Chatbot with ElevenLabs and OpenAI",
"nodes": 0,
"category": "AI & LLMs",
"status": "active",
"version": "1.0.0"
}Note: This is a sample preview. The full workflow JSON contains node configurations, credentials placeholders, and execution logic.
Get This Workflow
ID: a6d52c63-42a3...
About the Author
N8N_Community_Pick
Curator
Hand-picked high quality workflows from the global community.
Statistics
Verification Info
Related Workflows
Discover more workflows you might like
RAG Workflow for Stock Earnings Report Analysis
Analyze stock earnings reports using Retrieval Augmented Generation (RAG). This workflow leverages Pinecone for vector storage and Google Gemini for embeddings and chat, enabling detailed trend and outlier analysis from financial documents.
AI-Powered Conversational Agent with Tools
This n8n workflow creates an AI conversational agent that leverages multiple tools, including Wikipedia and a weather API, to answer complex user queries. It utilizes a buffer memory to maintain conversation context.
AI Agent to Chat with Your Search Console Data using OpenAI and PostgreSQL
Leverage an AI agent to query your Google Search Console data. This n8n workflow uses OpenAI for natural language processing and PostgreSQL for chat memory, allowing interactive data retrieval.