Booking.com Hotel Listing Scraper with AI Formatting - n8n Workflow

Market Researchers: Needing up-to-the-minute pricing and availability data from major travel sites.
Data Analysts: Requiring structured, pre-cleaned data feeds for competitive intelligence.
n8n Automation Specialists: Looking for advanced examples of integrating web scraping services (Bright Data) with conversational AI agents (Langchain/OpenRouter).
Developers: Seeking ready-to-use n8n templates for complex data pipelines.

This powerful n8n workflow provides a sophisticated solution for gathering competitive travel data. It addresses the challenges of dynamic web scraping and data presentation simultaneously. When triggered by a simple chat message specifying a city, the automation uses the Bright Data n8n node to initiate a targeted scraping job on Booking.com, ensuring reliable extraction of pricing, title, and address information. The architecture handles the asynchronous nature of web scraping by implementing a robust wait-and-check loop, ensuring the flow waits for the data to be fully collected. Once the data is retrieved, a key feature is the integration of an AI Agent (powered by OpenRouter and using the Langchain n8n node) and a Calculator tool. This combination processes the raw scraping output, cleans it up, performs necessary conversions (like showing prices in Euros), and transforms it into a highly readable, human-friendly list. This complete n8n template moves raw data directly to actionable insights.

This automation starts with a conversational n8n trigger:

1. Trigger & Input Capture: The process begins with the 'When chat message received' n8n trigger. The input (the target city name) is immediately captured and set as a variable using the 'parameters' Set n8n node.


2. Scraping Initiation: The 'Initiate batch extraction from URL' BrightData n8n node uses the city variable to launch a scraping job on Booking.com. This returns a snapshot ID.


3. Flow Control Loop: The workflow enters a monitoring loop. The 'Check the status of a batch extraction' n8n node polls Bright Data using the snapshot ID. The 'check if result ready' If n8n node validates if the status is 'ready'.


4. Waiting (If Not Ready): If the status is not ready, the flow is routed to the 'Wait' n8n node, pausing execution for a set duration before looping back to check the status again.


5. Data Download: Once the scraping is complete, the 'Download the snapshot content' BrightData n8n node retrieves the full JSON data set containing hotel listings.


6. Cleaning and Preparation: The raw data is cleaned using the 'clean data' Set n8n node, mapping important fields like title and prices. The 'Aggregate' n8n node then collects all individual hotel items into a single payload.


7. AI Transformation: The aggregated data is passed to the 'Human Friendly Results' Langchain Agent n8n node. This agent, configured with an OpenRouter Chat Model and linked to the 'Calculator' tool, follows system instructions to synthesize the data, convert currencies (using the Calculator), and output a beautifully formatted, conversational list.


8. Final Response: The resulting human-readable list is returned as the final output of this complete n8n workflow.

To deploy this powerful n8n template, follow these steps:

1. Import: Copy the provided n8n workflow JSON and import it into your n8n instance.


2. Bright Data Credentials: You will need valid credentials for the Bright Data n8n node. Create or select an existing Bright Data API key connection.


3. OpenRouter Credentials: Set up credentials for the OpenRouter service, required by the 'OpenRouter Chat Model2' n8n node. This typically involves providing your OpenRouter API key.


4. Trigger Setup: Ensure the 'When chat message received' n8n trigger is correctly linked to your chat interface (like Slack, Telegram, or a custom web hook) to accept user input (the city name).


5. Testing: Execute the n8n workflow to verify all connections and logic, especially the loop control and the AI processing stages.

This n8n workflow utilizes a complex blend of web scraping, control flow, and AI nodes:

When chat message received (Langchain Chat Trigger): The starting point, acting as the primary n8n trigger, initiating the automation upon receiving a user message containing the target city.
Set (parameters): Extracts the chatInput (the city name) from the trigger payload, preparing it for the scraping operation.
Initiate batch extraction from URL (BrightData n8n node): Starts the Booking.com web scraping job using the provided city name, returning a snapshot ID.
Check the status of a batch extraction (BrightData n8n node): Periodically checks the status of the scraping job.
If (check if result ready): A core n8n node for flow control, checking if the Bright Data snapshot status equals 'ready' before proceeding.
Wait: Implements a time delay if the scraping is still in progress, allowing the n8n workflow to be asynchronous and robust.
Download the snapshot content (BrightData n8n node): Retrieves the completed scraping results as raw JSON data.
Set (clean data): Standardizes the data structure by selecting and renaming key fields (title, address, prices) before AI processing.
Aggregate: Collects all individual hotel item records into a single array for efficient AI input.
Human Friendly Results (Langchain Agent n8n node): The centerpiece of the AI processing. This n8n node uses system instructions and the attached tools/LLM to transform raw tabular data into a conversational, well-formatted output.
OpenRouter Chat Model2 (n8n node): Provides the underlying large language model capabilities used by the Langchain Agent.
Calculator (Langchain Tool n8n node): Attached to the Agent, allowing the AI to perform mathematical operations, specifically converting scraped prices to Euros as instructed.