Covalent

Bonding Women in STEM

Comment

Inspiration

Women technologists make up a mere 28.8% of the tech workforce today. At this rate, it could take 12 years before women see equal representation in the industry. It is evident that there exists a significant diversity gap in STEM-based fields. Furthermore, our research indicates that mentorship from women in STEM increases female students' interest in pursuing a career in these fields (specifically when the mentorship takes place during grades K-12). Thus, we were inspired to explore ways in which technology can empower high school age women to pursue a career in STEM through mentorship, opportunities, and resources.

What it does

Covalent is a web app used to connect young females, specifically those in high school, interested in STEM to an experienced mentor in their fields of interests. Users are able to sign up as either a student or professional, where they are then able to make a profile. From there they are asked to choose up to three fields of interests in STEM as well as the type of mentor they prefer, hobbies/passions and their areas of strength. Through an algorithm designed to match students to the most appropriate mentor based on their selections; users are then matched to a mentor. After being matched, students then have the option to schedule a meet up with their matched mentor through a calendar.

How we built it

Covalent’s website was made using HTML/CSS and vanilla JavaScript. The mentor matching algorithm was written in Python and uses CSV files of mentor and mentee data as input. We also used Figma to design the layout of our website.

At first, we wrote the algorithm as a greedy algorithm that starts with a random mentee and finds the mentor with the closest overall match, including subject interest, hobbies, and skill sets. However, we wanted to ensure that as many mentees are paired with mentors in their subject(s) of interest as possible.

The new algorithm optimizes this by first creating a mentor set for each mentee that consists of all mentors in their subject area(s). Then, the algorithm pairs the mentees, starting with the mentee with the smallest mentor set. As pairs are matched, mentors are taken out of the remaining mentor sets.

If a mentee has multiple mentors in their set, they are matched with the closest matching mentor in terms of all their form responses. If a mentee has no remaining mentors in their set, the algorithm reverts to the greedy algorithm and finds the closest matching mentor overall from all remaining mentors.

Challenges we ran into

A challenge from the beginning of the Hackathon was working with teammates we had never met before. On top of this, having the Hackathon online, made it more difficult to communicate and work together. However, through communicating our strengths and skills from the start, we were able to produce a working web app.

Accomplishments that we're proud of

We’re proud of the following accomplishments:

-Pulling this project together remotely from 3 different time zones, without knowing each other beforehand

-Determining the mentor-mentee matching algorithm. We were stumped on how to prioritize some attributes over the other and still have optimal matches

-Translating our Figma designs and building out 6 fully functional user-interfaces on the front-end

What we learned

We’re excited to have learned the following:

-Some of us were beginner hackers, so it was energizing to build a product from end-to-end

-Developing a recommendation algorithm, optimizing its performance, and implementing it in Python

-Converting our beautiful Figma designs into responsive web pages

What's next for Covalent

Future steps for Covalent include the addition of an opportunities dashboard with local events near them, potential people they can connect with, as well as resources such as available scholarships and internship opportunities. Other improvements involve a meetup planning tool, safety features, and making the existing recommendation engine more robust.

Built With

Share this project:

Updates