Inspiration
Inspired by the statistics posted by the Harvard Business Review and the imposter syndrome that women in technological fields feel, we created Zera, a platform that helps women work on their self-doubt, build on their existing skills, and encourage them to apply to roles even if they aren't the "perfect" candidate.
What it does
Zera has a "Resume Match" feature that prompts the user to create a profile based on their background. After that users can upload their resume and a job listing they are interested in. The algorithm will compute a percentage of how well they fit the job listing. As long as it's 60% or above users are encouraged to apply for the job because according to the Harvard Business Review that has never stopped the male counterparts from doing so. If their match score is below 60% then the breakdown shows them what they matched with and what they can improve.
There's also a "Journaling Prompts" feature which encourages the user to write down and reflect on their emotions and doubts, where they come from, why they are feeling that way, and more, helping them work on their feelings around applying for jobs and improving their head space.
Lastly, we have a "Resume Quiz" feature where users can upload their resume and receive personalized practice questions to build your confidence when it comes to interviews and marketing themselves as an applicant.
How we built it
We used Java to program a backend component to the project that utilized the Scanner class to take user input. This allowed users to input information to create a profile. It takes resume and job description txt files and computes match scores (a percentage out of 100) between the two. The program currently tokenizes and cleans the words. It then creates a unique set of words between the two files and computes TF-IDF vectors. It then calculates cosine similarity to determine the match rate.
Challenges we ran into
We ran into challenges around working with Figma and created our wire frames. As it was a new platform for all of us, we had to learn new features and properties and implement user-friendly and accessible design while maintaining an aesthetically pleasing prototype. Additionally, we struggled with issues regarding github. Despite adding teammates as contributors they couldn’t access the repo or make changes. This forced us to have to work on the backend using the VS code liveshare feature, but we were unable to utilize that feature when we weren’t together, so post hackathon day 1 we had to work on separate files. Sadly, that resulted in a lot of bugs that needed to be fixed.
Accomplishments that we're proud of
As a team we are proud that we accomplished the core requirements of our program, which were the following: creating a program that allows users to make user profiles computing a match score using a user resume and job listing
What we learned
We need to familiarize ourselves more with is our tokenization. We are currently just using regular tokenization but we want to use an NLP to do this as it will be more accurate and increase efficiency.
What's next for Zera
As a future scope and development for Zera, we come up with the following features: Saving user info, so that users are able to log-in to the program instead of having to register every session. Getting the platform to host a portfolio of different companies. Then we can use the hobbies users incorporated in their portfolio to suggest job postings they can apply to based on the amenities offered at the company. Ex. if a user likes biking we’ll recommend them a job at Fidelity’s Merrimack campus
Log in or sign up for Devpost to join the conversation.