Inspiration Our team has always been interested in learning more about tools related to AI and when we saw that Geico had a challenge which asked us to make an AI tool to help their recruiting process, we decided to take on this challenge as we would learn something new.
What it does The ScreenAid tool helps recruiters screen resumes with features like keyword matching which includes a search for similar search, it filters out candidates by their GPAs and expected graduation date if necessary. The recruiter just has to put inputs of these fields, and the rest is done by the app.
How we built it We first built a PDFParser on Python which converts PDF into text. The app then matches the tech skills and soft skills inputs by the recruiters and takes out what percentage of the resume matches the input. It then compares the candidates' resumes' with the threshold put in by the recruiter and if it is less than the required GPA, it rejects the candidate.
Challenges we ran into Running Flask smoothly on our computers was a huge challenge. We had a lot of difficulty integrating the backend code into the front end code, which is the reason why we are currently not showing the output on our browser, but in a file output instead because of the limited time we had.
Accomplishments that we're proud of building a PDFParser with Python, Building a keyword match, and integrating similar word search with it with the help of Word2Net. We haven't used any Azure tools to power up our app and it still works like it should.
What we learned Building an AI tool is not as hard as it sounds.
Log in or sign up for Devpost to join the conversation.