LinkedIn Automation leverages AI tools to automate the creation of LinkedIn posts. It scrapes the web for the latest tech news and converts the information into engaging summaries. Then, it utilizes the LinkedIn API to share these summaries as new social media posts. We chatted with Daniel Rosel and Ismael Picazo, both pursuing a Bachelor in Computer Science & Artificial Intelligence, to learn more about their project.
What was the inspiration behind LinkedIn Automation?
With so much news coming out of the robotics and AI field, it’s hard to keep up with new research and stay informed on the latest progress. Our goal is to provide our followers with informative summaries without overwhelming anyone with unnecessary updates or irrelevant news.
Diving into the technical aspects of your project, what software and tools did you utilize?
We used Python for scripting and automation, along with various scraping technologies to gather data from the web. We also implemented proxy scraping techniques to manage proxies and avoid IP bans. For automation, we employed tools like Zapier and Linux scheduling software such as Cron. Additionally, we leveraged the OpenAI API for natural language processing; we’re currently moving toward self-hosted Large Language Models (LLMs) to enhance our text generation capabilities.
What challenges did you face during the project?
One significant challenge we encountered was managing the risk of IP bans while scraping data from various websites. To overcome this, we implemented proxy scraping techniques that allowed us to distribute requests across multiple IP addresses, reducing the likelihood of getting banned. That said, it’s important to note that LinkedIn Automation respects all the restrictions placed by websites that prohibit scraping.
Another problem related to the consistency and quality of the summary texts generated for LinkedIn posts. Initially, we relied solely on the OpenAI API for natural language processing, but the content generated sometimes lacked coherence or relevance. So, we started experimenting with self-hosted LLMs. This allowed us to fine-tune the models according to our specific needs, resulting in more accurate and contextually appropriate posts.
By integrating self-hosted LLMs with our existing automation tools and scheduling software, we significantly improved the quality and reliability of our posts. This shift also promises to cut costs associated with using third-party models. However, we’re still addressing issues with hallucinations to eliminate misinformation or discrepancies in our posts.
How did teamwork and collaboration contribute to your project’s success?
Teamwork was a pivotal component of this project. Our idea became a reality, thanks to the wide range of abilities contributed by members of the group. By working together, we could leverage each other’s unique skills to create something we couldn’t have individually. Once we figured out the best way to organize ourselves, we were able to ensure that everything fit together and worked well.
What key lessons or skills have you gained from working on this project?
One of the most important takeaways from this project was how to integrate different technologies. We had to implement several measures to connect the script with our self-hosted models, such as using Docker to ensure that the program could run in a stable manner independently of machines and tools.
Furthermore, building LinkedIn Automation provided invaluable experience with web scraping tools and API usage. We haven’t just enhanced our technical skills; we’ve learned how workflows and tasks can be automated using the latest innovations in AI.
Do you have any plans for further development?
LinkedIn Automation has a lot of potential as harnessing AI for content creation and social media engagement is a very relevant topic right now. In the future, it would be interesting to try scaling this idea to other social media platforms and club functions, such as emails and newsletters.
Tech innovation put to real-world use
LinkedIn Automation showcases the innovative spirit and technical prowess of students at IE University. By blending classroom learning with practical application in the Robotics & AI Lab, the Bachelor in Computer Science & Artificial Intelligence shapes professionals who can develop solutions for real-world challenges. This kind of project not only enhances students’ technical skills, it also prepares them to lead in the ever-evolving tech landscape.
