Skip to content

README.md

Latest commit

 

History

History
195 lines (143 loc) · 4.76 KB

README.md

File metadata and controls

195 lines (143 loc) · 4.76 KB

Python Requests dotenv

Extract LinkedIn profiles from Google SERP API

HasData_bannner

Companion code for a YouTube video showing how to find publicly visible LinkedIn profiles from Google search with the HasData Google SERP API.

Lightweight tutorial project: query Google SERP data, extract LinkedIn profiles, emails, location etc. From search results, and rank the best match for a person.

Watch the video

This repository includes small Python examples for:

  • Scraping LinkedIn profiles of people with the given Industry, Job Title and Location
  • Extracting info like full name, location, job title etc. From a given file of profiles with the help of AI.
  • Extracting the email of every person in a given file.
  • Extracting company information given a file that include a list of company names, With the help of AI

Quick Start

pip install -r requirements.txt

Create .env use LLM_SITE is only if you're using an LLM aggregator

HASDATA_API_KEY=your_api_key_here
LLM_KEY=your_llm_key_here
LLM_SITE=your_llm_site_here

Run the batch example:

python src/example_1.py

Workflow

Scrape LinkedIn profiles of people with Google SERP 
       |
       v
Inference those profiles with AI to extract names, location, followers, company they work on, etc. Automatically
       |
       v
From the AI extractions we can use the name and Google SERP api with the keyword "email" to enrich the data with emails
       |
       v
From the AI extracted company names we can use Google SERP api to get more info on the company.
       |
       v
Save all the information and all the steps in JSON and CSV files.

Project Structure

extract-emails-from-google-search/
|-- assets/
    |-- banner.png
    |-- youtube-preview.png
|-- output/
|-- src/
    |-- __init__.py 
    |-- api.py 
    |-- example_1.py 
    |-- example_2.py 
    |-- example_3.py 
    |-- example_4.py 
    |-- llm.py 
    |-- utils.py 
|-- .env
|-- .gitignore
|-- LICENSE
|-- README.md
|-- requirements.txt

Requirements

  • Python 3.10+
  • A HasData API key
  • A LLM API key

Install dependencies:

pip install -r requirements.txt

Configuration

Create .env in the project root, LLM_SITE is only if you're using an LLM aggregator

HASDATA_API_KEY=your_api_key_here
LLM_KEY=your_llm_key_here
LLM_SITE=your_llm_site_here

The scripts load this variable automatically with python-dotenv.

Scripts

src/example_1.py

The simplest example. It:

  1. uses HasData Google SERP to search LinkedIn profiles.
  2. scans 50 pages per query.
  3. outputs a file in output/

Run:

python src/example_1.py

src/example_2.py

Single-person matching mode. It:

  1. reads the data from output/n
  2. uses LLM to understand the search results
  3. extracts full name, company name, location, job title etc.
  4. outputs a file in output/

Run:

python src/example_2.py

src/example_3.py

Batch mode for multiple people from CSV. It:

  1. reads the data from output/
  2. uses HasData Google SERP api to search emails
  3. finds the best matching email with confidence score
  4. outputs a file with update info in output/

Run:

python src/example_3.py

src/example_4.py

Batch mode for multiple people from CSV. It:

  1. reads the data from output/
  2. searches the companies the LLM found from LinkedIn profiles
  3. using a LLM extracts company ceo, headquarters, industry involved, etc.
  4. outputs a file with update info in output/

Run:

python src/example_4.py

Notes

  • Results depend on what Google snippets expose at request time.
  • This approach only finds emails that appear publicly in search snippets.
  • Accuracy is limited when multiple people share similar names.
  • There's a small chance the LLM might hallucinate.
  • API usage depends on your HasData account and quota.

Why This Repo Exists

This project is meant to be extra material for a YouTube tutorial, so the code stays small, readable, and easy to follow. The focus is on demonstrating the core idea clearly rather than building a production-grade pipeline.

Use Cases

  • lead research demos
  • enrichment experiments
  • tutorial material for scraping and SERP parsing
  • lightweight prospecting workflows

License

This project is licensed under the MIT License. See LICENSE.