RateMyResume

Inspiration

As 2nd/3rd years in college, we wanted to have a platform where we can get quickly get resume critiques online. However in order to do this, we would need to manually anonymize our resume. Thus, we thought of trying to create a project that does that for us.

In addition to anonymizing the resume, we also wanted to create a large database of resumes from real people (as opposed to the existing databases of templates and fake resumes) that users can view at anytime to use for inspiration for their own resumes. This will hopefully foster a large community that supports each other by giving career advice.

What it does

RateMyResume is a web-app that allows users to upload a resume in PDF format. Then, we will automatically anonymize the resume by redacting any personal identifiable information (PII) on it. Then, the user will have the option to post them into a forum like website that other people can view, critique, and comment on.

When the user decides to post their resume, it will be added into a large database of resumes that can be viewed by anyone at any time.

How we built it

We built the project using Next.js for the frontend, and FastAPI to help with processing the pdfs with Python, and making API calls to other APIs like Amazon Comprehend to assist with utilizing Python in the backend, which simplified a lot of the code for us.

Challenges we ran into

Our biggest challenge was building the infrastructure of the web-app. As beginners, we didn't know a lot about which technologies / services were good to use for what we wanted to accomplish with our project. By talking to some mentors, we were able to gain some advice and be led in the right direction. For example, we didn't know that the Amazon Comprehend API even existed and what FastAPI was.

Connecting the database to the frontend and backend. Since we did not have that much experience with the backend, we were struggling a lot with our first usage of FastAPI, and how it connected the frontend and backend.

Displaying PDFs on the Next.js app as a clickable grid was also challenging.

Accomplishments that we're proud of

1) Integrated a disqus comment section for each resume
2) Used FastApi to process the pdfs with python and send them back to the Next.js app
3) Learning about various AWS services & integrating Comprehend into our project

What we learned

How to use FastAPI, and how to efficiently send post requests from a Next.js app to FastAPI in order to use python with Next.js. We also learned how to utilize the Amazon Comprehend API and call it when needed for our project.

What's next for RateMyResume

We would love to actually get a community started on RateMyResume by having people upload their resumes. We would also like to add a manual review stage before the user posts their resume, where the user can choose to redact additional information or restore redacted text. Though the NLP model used by the Comprehend API is usually quite accurate, the way the PyMuPDF library searches for the PII words is sometimes inaccurate, causing the black boxes to appear where we don't want them to.