Inspiration
Co:Here inspired us to also to use natural language processing. After bouncing ideas off of each other, we landed on the idea of creating a way to filter LinkedIn profiles for a specific internship, analyze the skills of those individuals using NLP, and output important skills and traits the company is looking for. This would be a very useful application for students currently looking for internships, as it can act as a guideline for what the student needs to practice and excel in.
What it does
As we take in data from resumes, we first grouped the student/resume with a specific job. This allows us to sum up the skills of all persons under the internship, understanding that whichever skills show up most frequent would most likely be _ THE _ desired skill for that company.
How we built it
We created our own database based on LinkedIn profiles (specifically the skills section) of 12 current CS students at Cal. Through Excel and csv files, we were able to move these results into tables, separating them based on job and then skill.
Challenges we ran into
Some challenges were that we realized that it would be challenging to obtain a LinkedIn API, in order to access a database of profiles that have experience with a certain company. LinkedIn's privacy rules would further prevent us from webscraping their profiles, as it is not possible to view certain profiles.
Accomplishments that we're proud of
Overall, we are proud of being able to reflect how the idea would have been implemented if we did have access to the API. Creating the beginnings of a website was also a great way to showcase the final, cleaned tables for each job showing the top 5 skills of the relevant internship/job. Although there is still much to be worked on, we are glad we were able to create something with the idea that we came up with.
What we learned
Although we were only here for a days, we learned about many topics such as APIs and companies such as Co:Here. We learned about natural language processing and discovered how tangible this aspect could become with its fluidity. We learned how to bridge the gap between coding and creating something with code. From the workshops, we learned how to make a webpage, which we used for our project. We learned that starting a project from not only zero starting code, but also little knowledge in full-stack developing is an incredibly arduous task. From the think-tank to lines of code, it was a rollercoaster. Every step of the way felt like we needed to move mountains, but we're glad we did not give up and participated still. Overall, CalHacks was an extremely rewarding experience, and has sparked inspiration in all of the team members to continue creating!
What's next for GetaJob
We still have much we want to add. We would like to add a search bar on the website so that a simple search would lend someone straight away to the end result. Yet this would mean much more work behind the scenes, through the LinkedIn API and webscraping, broadened through multiple and various iterations.
Log in or sign up for Devpost to join the conversation.