About our Project:
Let's be real...the job market is absolutely cooked right now.
Imagine this, or if you've experienced this... , you spend hours perfecting your resume, hit "apply" and then... rejection.
Here's the brutal truth: 75% of resumes don't even make it to a human recruiter. They're getting thrown into a digital trashcan by Applicant Tracking Systems (ATS), basically an application that scan your resume for keywords and formatting. And get this: 98% of Fortune 500 companies use these systems. So unless you got a lot of industry connections, you're going to deal with this.
And the worst part is, you could be perfectly qualified, overqualified even, for the job, but if your resume says "managed a team" instead of "led cross-functional initiatives" or whatever buzzword the ATS is hunting for, you get rejected.
That's where Un-"Unemployed" comes in.
Upload your resume to our platform, and our AI-powered model (shoutout to Gemini) analyzes it like a pro recruiter would—but without the judgmental side-eye. We identify weak spots, missing keywords, and formatting errors that make ATS reject you instantly.
But we don't just dump a wall of criticism on you and wish you luck. We give you a personalized to-do list that breaks down every single area for improvement. As you make each fix, you check it off the list. It's satisfying, it keeps you organized, and most importantly, it ensures you don't miss a single optimization before you hit that "submit application" button again.
Because honestly? You deserve to be un-unemployed.
How we built it
We built this using a three-layer architecture.
Frontend: Built with React, HTML, and CSS to create a clean user interface. Users upload resumes, view their personalized to-do list, and track improvements.
Backend: Powered by FastAPI. It handles file uploads, validates PDFs, and orchestrates the analysis pipeline. We chose FastAPI for its speed, built-in file upload support, and async capabilities.
AI Layer: Google's Gemini API does the heavy lifting. When you upload a resume, our backend sends it to Gemini with prompts that analyze it like an experienced recruiter: scanning for ATS red flags, missing keywords, weak verbs, and formatting issues. It returns actionable feedback packaged into a to-do list.
The Pipeline: User uploads → React captures file → Axios sends to FastAPI → Backend forwards to Gemini → AI analyzes → Backend packages JSON response → React displays results. We used CORS middleware so our frontend and backend could actually talk without issues.
Challenges we ran into
We struggled with merging the backend with the frontend, causing us to have to redo the CSS many times. In addition, for the frontend, we struggled with centering the text in the middle of the website. For the backend, we struggled with processing the user's input data which was in the form of a pdf. We found it hard to put it into Gemini to process it.
Log in or sign up for Devpost to join the conversation.