Being university students, all of us have gone through the process of applying to hundreds of jobs over the last few months. This has been a tedious and extensive process and we got to learn a lot along the way. Our project takes resumes and recommends tailor-made jobs for users, making their experience applying to jobs a lot better. In addition, it may help reduce the gap by providing access to resources to help low-income users improve their resumes and have a better chance of obtaining a job as compared to those that can afford external help.

What it does

Resumate is a web application that will make suggestions to improve your resume. It utilizes natural language processing (NLP) to analyze the text from your pdf resume and a Language Model from Cohere to give specific, personalized examples of what could make your resume pop to recruiters.

How we built it

Building it was a long and complex process that involved many stages. We divided the whole program into two major steps, first converting the PDF of the resume into a text format using optical character recognition (OCR), and then processing as required. By following these two major steps, we were able to build a program that recommended suggestions to build on the user’s resume. We wrapped it all up using Streamlit to make our data pretty, translating dirty Python backend to beautiful front-end.

Challenges we ran into

One challenge we ran into was finding a way to utilize Cohere in our project. We knew that we wanted to use it for our project, as it involved an NLP problem. There was a multitude of suggestions for the use of Cohere such as summarization and text classification. We managed to take full advantage of its Large Language Model to essentially type instructions directly and get an outputted recommendation for resumes. Through the careful revision of the docs and testing out code in the playground, we were able to apply Cohere to our own program in a useful way.

Accomplishments that we're proud of

Our proudest accomplishment for this hackathon was by far being able to utilize Cohere for the project in such a short amount of time. Understanding API usage and reading docs usually take a long time, but we were able to incorporate it into our code in under 24 hours. This was the last step of our plan, and being able to bring our idea to fruition was highly rewarding in itself.

What we learned

From understanding API to debugging numerous errors, we learned how Cohere can generate Al to create and organizes texts for various platforms. In addition, we will be able to expand our knowledge of python by completing different steps within the projects. Overall, we will be able to grasp a far greater knowledge of python and cohere, which will definitely benefit us in the future.

What's next for Resumate

In the future, we hope to expand Resumate to provide more additional support in creating effective resumes. These might include providing real-time job recommendations, updating the resume format, or creating a suitable cover letter to tailor the resume. As well, we hope to develop Resumate into a greater web application to serve a greater audience such as students, teachers, and web developers.

Built With

Share this project: