LinkedIn Profile Deep Scraping, AI Analysis, and Supabase Storage - n8n Workflow

Markdown-formatted text defining the ideal user:

Recruiters and Talent Sourcers: Automating the deep analysis of hundreds of candidates.
Sales and Business Intelligence Teams: Gathering comprehensive, structured prospect data.
n8n Automation Engineers: Users seeking advanced examples of combining web scraping, LLM processing (Google Gemini), and database operations within a single, highly efficient n8n workflow.
Data Engineers: Needing a reliable, scalable system to push web-scraped data into a managed SQL database like Supabase.

Leveraging publicly available data for intelligence gathering is crucial, but raw data requires structure and summarization. This comprehensive n8n workflow solves this by linking a powerful deep extraction tool (Bright Data) with state-of-the-art AI analysis (Google Gemini via n8n's Langchain nodes).

The automation begins with an n8n trigger (a webhook) that receives a target LinkedIn URL. It immediately scrapes the profile, then forks into five parallel AI processes to extract specific information (Basic Info, Emerging Roles, Skills, Summaries, and Detailed Experience). All these extracted data points are merged and aggregated before being stored in Supabase. This advanced n8n template demonstrates high-efficiency parallelism and robust database handling, checking if a record exists before deciding whether to create a new row or update an existing one, ensuring data integrity.

The entire process is initiated by an incoming Webhook, serving as the primary n8n trigger.

1. Trigger and Input Setting: The Webhook receives the LinkedIn profile URL. The subsequent Set the Input Fields n8n node prepares this URL for the scraping operation.


2. Web Scraping: The workflow utilizes the specialized Bright Data n8n node to Access and extract data from a specific URL, performing a deep scrape of the LinkedIn profile content.


3. Error Check: The flow immediately checks the scrape response status code. If the status is not 200, the If status code <> 200 n8n node routes the flow to Respond to Webhook with an error message.


4. Parallel AI Processing: If successful, the raw scraped data is passed to five separate branches, each using a specialized Langchain InformationExtractor n8n node powered by a dedicated Google Gemini Chat Model. This simultaneous processing extracts:

1. Data Consolidation: The Merge n8n node waits for all five parallel AI extraction processes to complete and combines their structured output. The Aggregate n8n node then formats the final, unified data payload.


2. Supabase Logic: The If force create? n8n node determines the database action. If creation is not forced, the flow attempts to Get a row in Supabase based on a unique identifier. The If record exist n8n node then dictates the final action: either Update Row (if found) or Create a row (if not found).


3. Final Response: The relevant Respond to Webhook n8n node sends a confirmation back to the initiating system, indicating whether the profile was successfully created or updated in the Supabase database.

To deploy this powerful n8n workflow, follow these steps:

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


2. Credentials Setup:

1. Webhook Activation: Activate the Webhook n8n trigger. Copy the test or production URL, as this will be the endpoint you call to initiate a profile scrape and analysis.


2. Testing: Execute the n8n template by sending a POST request to the webhook containing the target LinkedIn profile URL.

This advanced n8n workflow relies on several crucial nodes:

Webhook (n8n trigger): The starting point, configured to receive incoming POST requests containing the target URL.
Bright Data: Access and extract data from a specific URL: This specialized n8n node handles the complex task of scraping LinkedIn profile data securely and reliably, acting as the primary data source.
Langchain: lmChatGoogleGemini (Five Instances): These nodes provide the underlying large language model capabilities, enabling advanced contextual understanding and extraction based on the provided prompts.
Langchain: InformationExtractor (Five Instances): Key n8n nodes responsible for defining the specific structured output required (e.g., JSON schema for Basic Profile Info or Skills list) from the raw scraped text using the Gemini models.
Merge & Aggregate: Core n8n flow control nodes that combine the output of the five parallel AI branches into a single, cohesive item for insertion into the database.
If & Conditional Logic: The If status code <> 200, If force create?, and If record exist n8n nodes ensure robust handling of scraping failures and implement smart database logic (upsert/update or creation).

• Supabase: Get a row, Create a row, Update Row: These n8n nodes handle persistent storage, managing the database connection and performing the necessary SQL operations based on the automation's logic.