Resume Fit Analyzer

Landing page of the website
Page after uploading resume
Page after adding a job description link
Results after clicking "Analyze Fit"
Email entry location
Email with results that user can send to themselves

Inspiration

When we first entered the datathon, we got the privilege to attend a keynote speech by someone from Melissa Education. His primary discussion point was regarding our future job prospects as students who will graduate into a world increasingly governed by AI. This had us wondering about our own job prospects, and inspired us to create a project that would allow us to feel more prepared to enter this volatile workforce.

What it does

Our project allows users to upload a pdf of their resume as well as a link to a job/job description that they are interested in. In return, our trained model will classify the user as either "Not fit", "Potential fit", or "Good fit" for the specified job.

How we built it

We built this project using Pytorch to read data (which we found through HuggingFace) and train the classification model using a neural network. We then used FastAPI to create an API that would allow the model to pass its data/prediction to the frontend of our codebase, which was written using React.js. We additionally used Nodemailer to provide the option for users to send their results to their email.

Challenges we ran into

As a pair of individuals who entered this datathon having minimal experience and no clue of what to do, it was very difficult to learn how to train a model based off of preexisting data. This made finding a good dataset very difficult as well as the implementation of many features. The entire experience was a massive learning curve for us both.

Accomplishments that we're proud of

We are proud of the fact that we were able to create a full end-to-end project despite our extreme lack of experience. We were especially happy with how the presentation/frontend of the project turned out, as we didn't think it would look as aesthetically pleasing as it ended up looking.

What we learned

We learned how to train and test a model using a predefined dataset, how to create an API to allow a trained model to pass data to the frontend, and how to use Nodemailer.

What's next for Resume Fit Analyzer

In the future, we might implement an additional tool that provides tips for users to strengthen their resumes, as well as additional jobs that might fit their profile.