Inspiration
With an increasingly more competitive job market, applicants need a way to stand out from the crowd. One such way to do so is by tailoring and fine-tuning your resume to the job application's specific description and requirements. However, with the volume of job applications most applicants submit, it is difficult to maintain consistency while adjusting for each company's needs, which is where we got the inspiration for our project from.
What it does
This project is an AI-powered resume enhancement platform. It allows users to upload their resumes, typically in PDF format, and uses OpenAI's API to analyze, improve, and restructure the content into a professional and job-tailored format. The AI can correct grammar, enhance phrasing, and even suggest improvements based on job descriptions, which is provided by parsing the link to a job application. It returns the updated resume back to the user where they can either download or provide feedback on changes and send it back for revision with the user's prompt as the main point of revision.
How we built it
The user would input a resume into our web application, along with a link to the job they want to apply to, where it is then sent to our backend that's run on FastAPI,. It is then processed into a text-file and used as input to openAI's API where we prompt GPT-4 to convert the text-file into a strict JSON file based on a typescript template and make changes based on the parsed job description. This JSON file would then be compiled using a custom made compiler that converts the JSON into latex and returns the latex to the frontend as a pdf for the user to view
Challenges we ran into
File types were a major issue we ran into. Our project only supports pdf file types since multiple types file types have different ways of parsing their data and we cannot account for every single file type. Conversion between file types were also a big challenge, along with prompt engineering. It took a while to get GPT-4 to output the specific JSON that we wanted.
Accomplishments that we're proud of
Creating a live demo of real-time resume revisions and allowing the user to provide feedback on those revisions. Creating an endpoint that does the work of compiling the revised resume into latex which is then converted to pdf, creating a visually pleasing and nicely formatted resume.
What we learned
Utilizing FastAPI to create endpoints for backend Prompt engineering to get the most accurate response that suits our needs
What's next for Resumify
We want to make this more accessible, by that we mean we want to create a chrome extension. Here, we can automatically parse the job description and requirements along with the user's resume
Log in or sign up for Devpost to join the conversation.