Automated AI Company Profiling and Competitor Mapping - n8n Workflow

Market Researchers: Users needing automated, deep data extraction for competitive analysis.
Sales and Business Intelligence Teams: Professionals who require structured company data (mission, services, contacts) for lead enrichment.
n8n Automation Specialists: Users looking for advanced examples of integrating LangChain Agents, Mistral AI, and custom tools like Firecrawl and Tavily within an n8n workflow.
Data Engineers: Those building scalable data pipelines leveraging Supabase and Google Drive for archival.

This comprehensive n8n workflow template solves the challenge of converting unstructured website content into usable, structured company intelligence. It starts with an n8n trigger (a user form) to input a target website URL. The flow utilizes advanced AI techniques, specifically integrating Mistral AI with specialized scraping tools (Firecrawl for deep analysis, Crawlee for basic scraping) to ensure high-quality data extraction.

The core value lies in its duality: it profiles the company (extracting everything from SEO keywords and services to social media links) and then automatically launches a subsequent process using the Tavily search tool and Mistral to find and document direct competitors. All extracted data is seamlessly saved into dedicated tables within a Supabase database and archived as JSON files in Google Drive, making this a robust n8n solution for building a scalable B2B intelligence platform.

This powerful n8n workflow starts with the On form submission n8n trigger. Users provide the website URL and select the desired scraping method ('basic' or 'deep scraping').

1. Flow Selection: The Switch n8n node directs the data down one of two paths based on the user's scraping preference.


2. Deep Scraping (Firecrawl/AI Agent): If 'deep scraping' is selected, the MCP based scraper FIRECRAWLER n8n node (a LangChain Agent powered by Mistral and Firecrawl tools) visits the site and performs a highly detailed content extraction. A Structured Output Parser1 ensures the output strictly conforms to a detailed JSON schema, extracting 15+ data points (e.g., mission, vision, SEO keywords, social links).


3. Data Storage (Primary): The structured data is immediately saved to the companies table in Supabase. It is also converted to a JSON file and stored in Google Drive for backup.


4. Basic Scraping Path: If 'basic' is selected, the basic web scraper n8n node extracts the content, and the Information Extractor1 (also Mistral-powered) structures a less granular profile. This basic profile is stored in a separate company_basicprofiles Supabase table.


5. Competitor Analysis: Following successful data extraction and saving to Supabase, the company profile data (name, location) is fed into the find competitor n8n node. This agent utilizes the Mistral Cloud Chat Model3 and a specialized Web Search tool (Tavily search API) to identify direct competitors, using a dedicated Structured Output Parser to format the competitor data.


6. Final Storage: The competitor details are then recorded in the competitors table in Supabase via the Supabase4 n8n node, completing the profile enrichment. This highly efficient n8n workflow provides end-to-end intelligence.

To deploy this comprehensive n8n workflow template, follow these steps:

1. Import the n8n workflow: Copy the provided JSON and paste it into your n8n instance using the 'New' -> 'Import from JSON' function.


2. Credentials Setup: You will need to configure credentials for the following services used in the n8n nodes:

1. MCP Tools: Ensure the specialized n8n tools (Firecrawl and Tavily via MCP nodes) are correctly installed and accessible by the LangChain agents.


2. Table Preparation: Verify that the necessary tables (companies, competitors, companybasicprofiles, etc.) exist in your Supabase database with the expected schema fields to match the incoming structured JSON data.

On form submission (n8n trigger): Initiates the n8n workflow. Configured to accept the target Website URL and choice of scraping (basic or deep) and analysis options.
Switch (n8n node): Essential flow control. Routes the execution based on the user's selected 'Scrapping type' parameter.
MCP based scraper FIRECRAWLER (Agent): The core deep scraping component. This n8n node uses Mistral AI and Firecrawl tools to perform intelligent extraction of over a dozen structured data points from the website content.
Supabase (n8n node): Primary storage for the structured company profiles. It dynamically maps extracted JSON fields (like mission, services, social links) to the companies table.
find competitor (Agent): A secondary LangChain Agent that executes specialized searches using the connected Tavily API tool to identify market competitors based on the scraped company's name and location.
Structured Output Parser / Structured Output Parser1 (n8n node): Used to enforce a strict JSON structure on the raw text output generated by the Mistral agents, guaranteeing clean, database-ready data.

• Convert to File for storage / Google Drive2 (n8n node): Handles archival. The data is converted into a JSON file format and uploaded to a specified folder in Google Drive, ensuring redundant data retention.