AI-Driven Company Data Scraper via Conversational Chat - n8n Workflow

• Sales Development Representatives (SDRs) needing quick, verified company data.
• Market researchers and analysts who require structured data from corporate websites.
• Users looking for advanced examples of how to combine conversational n8n trigger nodes with powerful AI web scraping agents.
• Technical users who want to implement specialized, multi-step AI logic within an n8n workflow.

Manually sifting through corporate websites to find specific data points like CEO, address, or a company summary is time-consuming and prone to errors. This specialized n8n workflow solves this by setting up a fully conversational AI agent that acts as an expert web researcher.

When a user sends a URL via chat, the n8n trigger initiates a sequence where an AI agent validates the input, uses the powerful HttpRequestTool n8n node to navigate the corporate site (checking 'About Us' pages), extracts the required data, and ensures the output is translated into the requested language.

This robust n8n workflow structure minimizes hallucinations and guarantees structured, validated results, making it one of the most effective n8n templates for advanced data extraction tasks.

This n8n workflow starts with the When chat message received n8n trigger.

1. Configuration: The flow first initializes variables in the Config n8n node, defining the required extraction fields (e.g., CEO, Address, Summary) and the desired output language.


2. URL Extraction: The AI Agent (Extract URL) n8n node, powered by Google Gemini, analyzes the chat input to strictly extract the URL. This is critical for filtering out conversational noise.


3. Validation and Routing: The Check URL n8n node determines if a valid URL was found. If no URL is detected, the workflow prompts the user for input and loops back. If a URL is found, it proceeds to the main scraping sequence.


4. AI Extraction & Scraping: The AI Agent (Access URL) n8n node takes over. This agent is configured with an extensive system message acting as an expert web scraper. It uses the HttpRequestTool to fetch the HTML content, navigates to the 'Company Profile' or 'About Us' pages as mandated, and extracts the target fields. It strictly outputs the result in a CSV format, including a mandatory autoFetchStatus and autoFetchReason.


5. Data Processing: The ParsedCsv n8n node then converts the structured CSV output from the AI into standard n8n JSON objects.


6. Response Formatting: The ChatResponse n8n node formats the extracted company data into a clean, human-readable Markdown format, suitable for the chat interface.


7. Conversational Loop: Finally, the Respond to Chat n8n node sends the results, and the Request Next URL n8n node prompts the user for another input, closing the loop and keeping the conversational n8n workflow alive.

To deploy this comprehensive n8n workflow, follow these steps:

1. Import the Workflow: Copy the entire JSON code and paste it directly into your n8n instance using the 'New' -> 'Import from JSON' function.


2. Set up Gemini Credentials: This n8n workflow requires credentials for Google Gemini (formerly Google Palm/Google AI). You must configure the API key in both Google Gemini Chat Model and Google Gemini Chat Model1 n8n nodes.

1. Activate the Trigger: The When chat message received n8n trigger is set as a LangChain Chat trigger. Save the n8n workflow, ensuring it is marked as 'Active' in the top-right corner.


2. Customization (Optional): Review the Config n8n node to modify targetCompanyFields (e.g., 'Revenue, EmployeeCount') or change the language for localized results. This flexibility is a key feature of this n8n template.

When chat message received (n8n trigger): The starting point for this conversational n8n workflow. It waits for incoming chat messages to begin the process.
Config (Set n8n node): Defines initial variables like targetCompanyFields ('CEO,Address,Summary') and language ('English'). These variables dynamically control the AI agent's extraction behavior.
AI Agent (Extract URL) (n8n node): A specialized LangChain Agent responsible for isolating a valid website URL from the user's input, returning a strict JSON object. It utilizes the Gemini language model.
Check URL (IF n8n node): Performs core logic flow control, checking if the AI successfully extracted a URL. If not, the n8n workflow prompts the user for corrected input.
Google Gemini Chat Model (n8n node): Provides the necessary LLM connectivity for the AI Agents.
AI Agent (Access URL) (n8n node): The core extraction mechanism. It uses a detailed system prompt to enforce rigorous validation, navigation, and extraction protocols, guaranteeing high-quality, verified data. It is integrated with the HttpRequestTool n8n node.
HttpRequestTool (n8n node): An essential tool provided to the AI Agent, allowing it to browse the internet, fetch HTML content, and follow links, simulating a human web researcher.
ParsedCsv (Code n8n node): Manages the conversion of the complex, quote-enclosed CSV output from the AI Agent into usable n8n JSON data items.

• ChatResponse & Respond to Chat (n8n nodes): These nodes format the final output and communicate the results back to the user via the conversational channel, concluding this specific step of the n8n workflow.