ConnectHer

Logo
Homepage
Resume Upload
Start Swiping!
Job Matchmaking
Find Networking Events
HerSpace Discussion Forum
DMs

Inspiration

A Harvard study found that a lot of women don’t apply for jobs unless they meet every single qualification. It’s not just about confidence, it’s that many women see job requirements as strict rules, not suggestions. They often assume that applying without checking all the boxes would be a waste of time or even disrespectful. But that mindset can hold people back. The truth is, companies might actually be missing out on amazing talent because of how they write job listings. It’s time we start seeing job descriptions as guidelines, not gatekeepers. Introducing ConnectHer, inspired by the #1 dating app, Tinder, our goal is to make a job matching application style in efforts of making the process more casual, boosts confidence, and community driven. Using a resume upload, our app will take the user’s resume and parse the information in the model to find similarity of skills and experiences in the job description to personalize data to the user. There are other features including a map to locate jobs nearby that are recruiting employees as well as a discussion post page that the targeted audience can go to look for support within this uplifting community. There is a chat feature where the user can reach out to recruiters (vice versa) and other users without the need to post on the public discussion forum.

What it does

Job searching app that allows the targeted user to upload their resume and the app will generate jobs based on their experiences and skills. If no matches occur, suggestions are generated for the user, this includes steps moving forward, project ideas, and resume suggestions. There is a percentage “match” feature that is essentially your qualifications based on the information your resume uploaded and the job postings similarity. Qualifications will only show 50% or higher and a suggestion feature can be found to increase your qualifications score. At the bottom there is a navigation tool that takes you to our map feature as well as our community forum (HerSpace). The map allows the user to look at local networking opportunities, like conferences and local career fairs. It’s a great way to start small and build experience with potential local companies. HerSpace is our community forum where users can share questions, concerns, or uplifting messages. It’s designed to be a safe, supportive space for women to talk about the hiring process, job-related challenges, and everything in between.

How we built it

In Phase 1, we created and launched a minimum viable product (MVP) using Figma. We created a text file containing key skills frequently listed in tech-related job postings. We used the PyPDF2 library and regex (re) module in Python to extract text from PDF files. We used regex to parse key skills from the extracted text and return the set of matched skills.

Additionally, we wanted to be able to have a broader profile creation algorithm, so that when resumes are uploaded, less specialized but still relevant skills would be included in user profiles. To do this, we utilized a natural language processing (NLP) library known as SpaCy. After reading a PDF file, the program strips it into pure text, and using the SpaCy and KeyBert libraries, searches for keywords, such as relevant nouns, verbs, etc. and adds these to a .txt file.

In Phase 2, we plan to implement a system that stores uploaded resumes. When the user uploads their resume, this will be uploaded to AWS S3 cloud services. The SQL database will store the reference to the resume and associated user information so data can be easily retrieved when the user returns to the app. We also plan to implement the job suggestion algorithm for swiping. Matched skill percentages will be calculated and jobs with higher percentages will be suggested earlier.

Challenges we ran into

We tried utilizing an existing resume parser on GitHub, but there were issues getting it to run due to the fact that the repo was several years old and wasn’t being maintained anymore. It was outdated and this became a compatibility issue. So, we pivoted and built a simpler solution: a custom resume parser that compares skills from a text file to the extracted text from a PDF resume. We also realized we could not integrate the technical features into Figma, so we decided to showcase these features separately in the demo video.

Accomplishments that we're proud of

We are proud of the UI/UX aspect of our app and creating a visual for all of the features we offer. We are also proud of creating a resume parser that filters for specific skills and general key phrases.

What we learned

We learned how to utilize Figma to create a prototype of our app and also integrate new Python libraries. We learned how to use Natural Language Processing to search and categorize specific parameters to create a personal profile based on resumes. Overall, working in a team and clearly communicating our strengths and weaknesses; being put in a team of individuals we didn’t know before was a new challenge, but we overcame it quickly and learned to utilize each of our skills to create an innovative project.

What's next for ConnectHer

We plan to continue working on this project and bring Phase 2 to life. We plan to add a library to detect similar skills based on the text file containing frequently key skills. After, we would like to implement a database to store the user’s extracted resume data as well as cloud service like AWS past resumes. We hope to continue developing this project into a full website so it can be accessible across platforms in the future. Accessibility matters, especially in today’s world, and we want to make sure ConnectHer is easy to use and available to women looking for a close-knit community.