An Automated LinkedIn post generator was developed to create custom content for the most trending events of the day.
The Trending Post Generator obtains the 'Top Google Searches' for the day and combines them with prompts sent to Chat-GPT to produce the ideal LinkedIn post.
The entire process is effortless, fast, and results in highly engaging content for the LinkedIn audience.
The application employs a WebDriver instance with Selenium to access Google's daily trends, located at: https://trends.google.co.il/trends/trendingsearches/daily?geo=IL&hl=iw
For each trend-item,
it verifies if the search_count exceeds the value defined as
MIN_SEARCH_COUNT (specified in the /helpers/selenium_helper.py file).
If the condition is met, the application proceeds to access the associated article, extracts its content, and stores it in a dictionary within a search_results List, specifically linked to the respective trend-item.
Next,
the application iterates through the search_results items
and creates prompts by combining a basic-query text from the
base_linkedin_post_prompt.txt file with the
saved article content corresponding to each trend-item.
Subsequently,
the application sends the generated prompt to Chat-GPT,
utilizing the OpenAI API,
and saves the resulting conversation's answer in a .txt file
located in the /linkedin-posts directory.
Create a .env file.
Openai Api Key (Required):
save your secret key in
OPENAI_API_KEY variable.
Linkedin Login Credentials (Required):
save your email and password in
LINKEDIN_EMAIL and LINKEDIN_PASSWORD variables.
Your .env file should look like this:
OPENAI_API_KEY=kfdgjlkvldnvkljfdkjdf
LINKEDIN_EMAIL=mymail@gmail.com
LINKEDIN_PASSWORD=secretpassword
Change Prompt:
You can set your own default base prompt
inside the base_linkedin_post_prompt.txt file
Change Minimum Search Count:
Set a different value to get a better results
by changing the MIN_SERACH_COUNT value
in the /helpers/selenium_helper.py file
- Selenium
- Openai Api