As a business owner, I'm acutely aware of the importance of forging personal connections with our users, particularly in the nascent stages of our venture. However, the task of manually scouring LinkedIn for each user's profile is daunting and time-intensive. That's where Moonlit comes into play.
At this pivotal moment in our company's journey, we cherish the opportunity to personally engage with you, our valued users. This interaction is crucial for garnering feedback and rapidly refining your experience with Moonlit. The traditional method involved a tedious process: sifting through user data like emails, names, and project titles, then deducing the user's employer to pinpoint their LinkedIn profile. While this approach yielded reasonably accurate results, it was far from efficient. Hence, I decided to leverage Moonlit's capabilities to automate this process, which I will now outline (perhaps you're even one of the users who received a LinkedIn invitation from me before reading this!).
To begin, I created three text inputs within the Moonlit app corresponding to the user's name, email, and project name. The next step involved employing a Large Language Model (LLM) to utilize these dynamic inputs for formulating an effective Google search query. The LLM also needed access to Google's Search functionality to execute the search, analyze the results, and ultimately return the LinkedIn URL that best matched the user's details. To refine this process further, I provided additional context about Moonlit's typical user base. This context helps the tool differentiate between profiles, especially useful when encountering common names, prioritizing profiles with keywords like "Generative AI," "SEO," or "Content Marketing."
The results were fantastic! 🎉 Initially, inputting user data individually seemed more cumbersome than not using the application. However, Moonlit's true potential is unleashed when processing data in bulk. The application allows for the uploading of a CSV file, where each record corresponds to an input. This feature enables parallel processing at scale, efficiently fetching LinkedIn profiles for a multitude of users.
To initiate a large-scale operation, I uploaded user data and aligned the app's inputs with the respective columns in the data. Upon creating this Bulk Run Job, the entries are populated into a table. Running this job appends a new column to our original user data, indicating the LinkedIn profiles.
Admittedly, this method isn't flawless. Some users may not have a LinkedIn profile, or they might use different names online. Despite these challenges and the occasional need for manual verification, this approach has significantly reduced the time spent on this task. In the future, we could explore adding features like a ChatGPT-generated accuracy score for each profile, or further refining the search prompts for even better results.
I'm curious to know if you've employed Moonlit for any unique tasks or if you have any innovative ideas for its use. Your input is invaluable, and I'd be thrilled to hear your thoughts! 😊
Mohammad is a full-stack developer, and the founder of Moonlit Platform. He holds a Bachelor's degree in Computer Science & Artificial Intelligence, and is committed to continuous learning and skill enhancement. His journey is marked by a steadfast dedication to developing and delivering exceptional product experiences.