AI Agent: Logo Sheet Extractor & Cataloguer for Airtable

Unlock Automated Market Intelligence with this AI Agent

This n8n workflow deploys an AI Agent to transform visual market landscapes and competitor logo sheets into structured, actionable data within your Airtable base. Simply upload an image containing multiple logos (e.g., industry maps, comparison charts), and the AI Agent, powered by OpenAI's vision capabilities and Langchain, gets to work.

It intelligently identifies software, tools, or products depicted, extracts key attributes (like features or categories) from the visual context, and even determines similar tools shown. This information is then meticulously organized and upserted into your Airtable, creating a dynamic and growing knowledge base with minimal manual effort.

Key Features & Benefits

• AI-Driven Image Analysis: Leverages an AI Agent (Langchain with OpenAI's GPT-4o model) to 'read' and interpret complex images containing multiple logos and contextual information.
• Intelligent Data Extraction: Accurately identifies tool names, extracts relevant attributes, and infers relationships (like similar tools) directly from the uploaded image.
• Automated Airtable Population: Seamlessly creates or updates records in designated Airtable tables for 'Tools' and 'Attributes', including establishing relationships between them.
• Structured Output: Converts unstructured visual data into a clean, JSON-formatted list of tools with their names, attributes, and similar counterparts, ready for database insertion.
• Duplicate Management: Employs hashing for tool names to ensure data integrity and prevent redundant entries in Airtable during upsert operations.
• Streamlined Input: A simple n8n form trigger allows easy uploading of logo sheet images to kickstart the automation.
• Customizable Logic: The AI Agent's core prompt can be adjusted to fine-tune the extraction process based on the specific types of logo sheets or information you're targeting.

• An n8n instance (Cloud or self-hosted).
• OpenAI API Key with access to a vision-capable model (e.g., gpt-4o is used in the template).
• Airtable account and a Personal Access Token.
• An Airtable Base pre-configured with two tables: 'Tools' and 'Attributes'. Required fields:
  • 'Tools' table: Name (singleLineText), Attributes (multipleRecordLinks to 'Attributes' table), Hash (singleLineText), Similar (multipleRecordLinks to 'Tools' table).
  • 'Attributes' table: Name (singleLineText), Tools (multipleRecordLinks to 'Tools' table).

1. Download the n8n workflow JSON file.
2. Import the workflow into your n8n instance.
3. Configure your Airtable credentials: In each Airtable node, create or select your Airtable Personal Access Token credentials.
4. Configure Airtable nodes: For each Airtable node, select your Base ID and the correct Table ID for 'Tools' or 'Attributes' as appropriate.
5. Configure OpenAI node: In the 'gpt-4o' Langchain LLM node (or your chosen LLM node), enter your OpenAI API Key.
6. Review AI Prompt: Check the 'Retrieve and Parser Agent' node's systemMessage parameter. Adjust the prompt if needed to better suit the types of images or data you'll be processing.
7. Activate the workflow.
8. The workflow is triggered by a form submission. Open the form URL (typically https://[your-n8n-domain]/form/logo-sheet-feeder) to upload a logo sheet image and an optional additional prompt.