Inspiration
Due to this hackathon's focus on social good/justice, we started to brainstorm projects that could combine AI and social justice. This track of thinking quickly reminded us of Joy Buolamwini's book, Unmasking AI: My Mission to Protect What Is Human in a World of Machines. That book covers a topic called the "coded gaze," basically explaining that AI is not neutral, and in fact, pretty biased. This led us to make our current project in an attempt to combat that AI bias and thus social injustice.
What it does
ResumeLens takes in an input, as either text or PDF, to calculate the percentage of the input being biased against marginalized groups. With this score, ResumeLens will explain why the user's input is biased and suggest how the user should word their resume afterwards.
How we built it
We built ResumeLens with Figma, Next.js frontend, FastAPI backend, and a HuggingFace model to handle training for bias detection.
Challenges we ran into
We originally wanted to use a more direct approach to solving this issue of AI bias by creating a more bias-free resume scanner that employers could use. Still, we quickly pivoted once we realized that, while more effective in preventing the issue of bias, it may not be the best course of action overall. The issue with this idea is that the employer themselves would have to go out of their way to utilize this less biased program. There is no guarantee a company would use it, or even want to use it, in fact. The only way to guarantee an “even playing field” would be to have it implemented on the job seeker's side rather than the employer's. Additionally, we had hoped that we would be able to generate unique suggestions for each resume that could be used to adjust them; however, due to Char being the only one who knows how to code, and us running out of AI code, we sadly could not implement this.
Accomplishments that we're proud of
We are quite proud of the entire project, as it was my (Faaris) first-ever hackathon, and I came into this event with zero knowledge of any sort of coding or AI. I am especially proud of learning how to use Figma to create the frontend of our project. It was also Chars's first time using ML, so she is quite proud of our AI model.
What we learned
We learned how to utilize Figma and ML with a HuggingFace model.
What's next for Resume Lens
We honestly hope to continue to refine our design and eventually present it to a couple of student startup courses/events. Hopefully, we can eventually turn our project into a fully working website that can aid everyone in gaining an equal footing in the job market.
Log in or sign up for Devpost to join the conversation.