After brainstorming about the various 4 themes, we decided to focus on inequality within the job industry since this would bring more opportunities to a more diverse workforce.
What it does
This web application takes in Resume and Recommendation Letter files and parses through the file to extract the text. Our website then returns a modified version of the text with gender-neutral vocabulary(e.g. she/he is replaced with they) and initialized names(e.g. John Smith becomes JS) to reduce any biases during the selection process of jobs. The employer can then search our database with an applicant's initials and unique 5-character ID in order to find the applicant's contact information and real name.
How we built it
Our base was Python and we used the OCR API in order to extract text from the files. HTML/CSS and Bootstrap were used to create the website.
Challenges we ran into
OCR sometimes did not return the correct text, which broke our algorithm sometimes.
Accomplishments that we're proud of
Getting a simple version of this to work under 24 hours!
What we learned
OCR is very unreliable.
What's next for Blind re•cruit
Improving the algorithm to be faster and more efficient.