AI Social Media Link Extractor Agent for Automated Data Enrichment

Unlock Automated Lead Enrichment with this AI Agent

This AI Agent is designed to supercharge your data collection efforts. It autonomously crawls company websites, intelligently identifies social media profile links (like LinkedIn, X, Instagram, etc.), and extracts them into a structured JSON format. Think of it as a diligent research assistant that works 24/7 to keep your company data fresh and comprehensive.

This AI-driven automation streamlines the tedious process of manually searching for social media information, freeing up your team for high-value tasks. It leverages OpenAI's GPT-4o model combined with Langchain agent capabilities, equipped with specialized tools for web text and URL retrieval, ensuring robust and accurate data gathering from target websites.

Key Features & Benefits

• Autonomous Web Crawling: Navigates websites (e.g., sourced from your Supabase database) to find relevant information.
• AI-Powered Link Identification: Uses OpenAI GPT-4o to intelligently recognize and extract various social media profile URLs based on patterns and context.
• Specialized Langchain Tools: Employs dedicated sub-workflows (tools) for effective web interaction:
  • text_retrieval_tool: Fetches and processes textual content from web pages.
  • url_retrieval_tool: Discovers and lists all hyperlinks on a page, enabling multi-page crawling.
• Structured Data Output: Delivers findings in a clean, parseable JSON format (customizable schema), ready for integration into your CRM, databases, or other systems.
• Scalable Data Enrichment: Efficiently processes lists of companies to enrich your records with valuable social media touchpoints.
• Time Savings & Efficiency: Drastically reduces manual research time, allowing your team to focus on outreach, strategy, and other high-impact activities.
• Customizable & Extensible: The AI's behavior (via system prompt in the Langchain Agent node) and the desired output schema (in the JSON Parser node) can be tailored to specific needs or to extract other types of data.

• An n8n instance (Cloud or self-hosted).
• OpenAI API Key with access to a suitable model (e.g., gpt-4o is used in this template).
• Supabase account, project, and API credentials for database interaction.
• A Supabase table (e.g., 'companies_input') with company names and their corresponding website URLs.
• A Supabase table (e.g., 'companies_output') prepared to store the company name, website, and extracted social media links.
• Ensure target websites are publicly accessible for crawling; consider proxy configurations for extensive crawling.

1. Download the n8n workflow JSON file.
2. Import the workflow into your n8n instance.
3. Configure the 'OpenAI Chat Model' node: Enter your OpenAI API Key in the credentials section.
4. Configure the 'Get companies' (Supabase) node:
   • Select or create your Supabase API credentials.
   • Choose the correct Supabase table (e.g., companies_input) that contains company names and websites.
   • Ensure your input table has columns named name and website. If not, adjust the 'Select company name and website' node to map your column names accordingly.
5. Configure the 'Insert new row' (Supabase) node:
   • Select or create your Supabase API credentials.
   • Choose the target Supabase table (e.g., companies_output) where the extracted data will be stored.
   • Ensure the 'Data to Send' option is set to 'Auto-Map Input Data' or manually map fields if your target table has different column names for company_name, company_website, and social_media.
6. Review the 'Crawl website' (Langchain Agent) node:
   • The Text parameter defines the initial prompt to the agent, using the website field from the input.
   • The System Message guides the AI's overall behavior. You can customize this to refine its task or extraction focus.
7. Review the 'JSON Parser' node:
   • The Input Schema defines the expected JSON structure for the social media links. Modify this if you change the AI's output instructions or need a different format.
8. Test the workflow with a single company first to ensure correct configuration and data extraction.
9. Activate the workflow. It can be run manually or scheduled for regular data enrichment.