AI Knowledge Agent: OpenAI Assistant with Google Drive Document Integration

Unlock On-Demand, Document-Specific Answers with this AI Agent

This n8n workflow empowers you to build and deploy an AI Knowledge Assistant. This agent uses OpenAI's Assistant capabilities, fueled by a specific document from your Google Drive, to answer user queries. It's a two-part process: first, a setup flow creates the OpenAI Assistant, downloads your chosen Google Drive document (e.g., a PDF), uploads it to OpenAI, and links it as the sole knowledge source. Second, a chat interface allows users to interact with this highly specialized assistant, which will only use the provided document to formulate its responses.

Key Features & Benefits

• AI-Powered Document Interaction: Creates an OpenAI Assistant that deeply understands and answers questions based on a single document from your Google Drive.
• Google Drive Document Retrieval: Automatically fetches your specified document (supports conversion from Google Docs to PDF) to serve as the AI's knowledge base.
• Precision Answers (RAG-like): The AI Assistant is strictly instructed to use only the content of the uploaded document, ensuring factual accuracy tied to your source material and minimizing irrelevant information, similar to a Retrieval Augmented Generation approach.
• Interactive Chat Interface: Users can engage with the AI Knowledge Assistant through a built-in n8n chat trigger, getting instant answers.
• Customizable AI Persona: Define the name, description, and detailed operational instructions for the OpenAI Assistant to match your brand's voice and specific use-case requirements.
• Automated Assistant Configuration: The workflow automates the creation of the assistant, file upload to OpenAI, and linking the file to the assistant.
• Conversation Memory: Includes a window buffer memory to maintain context during chat sessions.

• An n8n instance (Cloud or self-hosted).
• OpenAI API Key with access to the Assistants API and a suitable model (e.g., gpt-4o-mini, gpt-4o, gpt-4). Note: The workflow example uses gpt-4o-mini.
• Google Drive credentials (OAuth2) configured in n8n.
• A document in Google Drive to be used as the knowledge source (the workflow example uses a Google Doc which is then converted to PDF by the Google Drive node).

1. Download the n8n workflow JSON file.
2. Import the workflow into your n8n instance.
3. Configure Credentials & Core Settings: a. OpenAI Nodes: Select your OpenAI API credential in all OpenAI-related nodes (typically named 'OpenAI', 'OpenAI1', 'OpenAI2', 'OpenAI Assistent'). b. Google Drive Node: Select your Google Drive credential.
4. Customize Your AI Assistant (Initial Setup - Run Manually): a. In the first 'OpenAI' node (this node creates the assistant, possibly named like '"Travel with us" Assistant' in the template): Customize the Name, Model ID (e.g., gpt-4o-mini), Description, and critically, the Instructions. Tailor these instructions to define how your AI should behave and its knowledge scope (e.g., 'You are a helpful assistant. Only use the uploaded file for answers. Do not use any other knowledge.'). b. In the 'Google Drive' node: Specify the File ID of the document you want the AI to use. If it's a Google Doc, ensure 'Google Docs to PDF' conversion is active under 'Options' > 'Google File Conversion', as the workflow uploads a PDF to OpenAI. c. Manually run the workflow once using the 'When clicking ‘Test workflow’' trigger. This performs the initial setup: creates the OpenAI Assistant, downloads your document from Google Drive, uploads it to OpenAI, and links the document to your Assistant. The subsequent nodes are designed to pass the necessary IDs automatically.
5. Verify Assistant Linking for Chat: a. In the 'OpenAI Assistent' node (this node handles the chat interaction): Ensure the Assistant ID field correctly references the assistant created in step 4c. This is typically handled by n8n's dynamic referencing if the nodes are connected properly. If needed, select the created assistant from the list after it has been generated.
6. Activate & Test Chat: a. Activate the workflow. b. Use the 'When chat message received' trigger to interact with your AI Knowledge Agent. Send messages via the chat interface that appears when you test this trigger or when you integrate it into a frontend.