ResumeAtlas

A privacy-first career app with a right to be forgotten. It visualizes admin-curated, anonymous resume data to show you what works. A Gemini AI then writes your personal action plan. Built solo.

Comment

Inspiration 🚀

The inspiration for this project came from a feeling every student knows all too well: the anxiety of the "black box." We spend countless hours perfecting our resumes, sending them into the void, and hoping for the best. I've watched my friends and myself agonize over the same questions: What skills does Google actually want for this role? Is my project good enough for a company like Wells Fargo? Am I even on the right track?

This is "resume roulette," and it's a lonely, inefficient process. I was inspired to build a solution that turns this isolation into collaboration. I envisioned a platform built for students, by students, that could safely and anonymously pool our collective career intelligence. I wanted to create a pane of glass to look through that black box—a living, breathing map that shows us the pathways our peers have successfully taken, empowering us all to navigate our own careers with data and confidence.

What I Learned 🧠

This hackathon was a steep learning curve, and I've grown immensely, both as a developer and a designer.

Technically, this was my first deep dive into building a full-stack application around a major AI model. Integrating the Gemini API was more than just a simple API call; I learned the art of context-aware prompting. I had to architect a system that could package a user's selection—company, role, top skills, sample projects—into a rich prompt to get a truly insightful, personalized summary back from the AI. It taught me that the power of an LLM isn't just in the model itself, but in the quality of the conversation you have with it.

Conceptually, my biggest lesson was this: for a platform built on user data, trust is the most important feature. My initial focus was on the visualization and AI. But I quickly realized that without a system to guarantee data quality and user privacy, the platform would fail. This led me to develop the admin curation queue and the "right to be forgotten" feature. I learned that building robust, ethical systems is just as critical as building cool features.

How I Built It 🛠️

This was a solo project, so I had to be the architect, developer, and designer all at once. I built it as a full-stack web application with three core layers.

The Frontend: I used React for a modern, component-based structure and TailwindCSS for rapid styling. The heart of the frontend is the interactive bubble chart, which I built using D3.js. Getting D3 to work seamlessly within a React environment to handle dynamic data rendering was a key technical challenge.

The Backend & AI Core: I chose a lightweight Flask backend to handle the core logic. Its main jobs are to process resume uploads and manage communication with the AI. For the AI, I used the Base44 integration to call the Google Gemini API. This is the "insight engine" of the app.

The Trust & Safety Layer: This is the most critical part of the backend. When a resume is uploaded, it's immediately anonymized. The extracted, non-identifiable attributes (skills, roles, project snippets) are not fed directly into the live data. Instead, they enter an admin curation queue. From a secure dashboard, an administrator can review, edit, and approve this data, ensuring no malicious or irrelevant information pollutes the system. This human-in-the-loop process guarantees data integrity. I also built the "right to be forgotten" pipeline, allowing users to request the deletion of their contributed data.

Finally, I designed and integrated our mascot, Brady, as a friendly guide to make the experience more welcoming and engaging.

Challenges I Ran Into 🧗

Building this alone in such a short time was my biggest challenge. The sheer scope—frontend, backend, AI, and a robust data curation system—was a true race against the clock. Managing my time and resisting the urge to add too many features was a constant battle.

My second major challenge was solving the "garbage in, garbage out" problem. How could I let users contribute data without letting the platform get overrun by spam or irrelevant information? A simple filter wouldn't be enough. The solution was the admin curation system. It was more complex to build, but it was the right way to ensure the platform's long-term value and trustworthiness.

Finally, making the AI truly useful was a challenge. A generic summary wouldn't cut it. It took several iterations of prompt engineering to get the Gemini API to produce the kind of concise, actionable, and personalized advice that a student would actually find helpful. It was a challenge of turning a powerful technology into a practical tool.

Current Status & Future Vision: Live and Growing 🚀 This project is more than just a hackathon prototype; it's a live application already making an impact.

Live Deployment (As of September 28th)

The platform is currently hosted and live on Base44, launching with a foundational database of 300 anonymized resumes. As of tonight, it is being deployed to its first user base: the members and officers of the ACM chapter at ASU. I am serving as the initial administrator, with a plan already in motion to grant curation access to club leadership. This will empower them to manage the data quality and ensure the platform's integrity as it grows.

Immediate Goal: Building the Data Foundation

With the platform now live for ACM members, and a database with 300 anonymized resumes, the immediate goal is to run our first "Resume Seeding Drive." By onboarding this initial cohort of users, we will build the foundational dataset that is crucial for providing meaningful insights. This marks the first turn of the project's "flywheel," transforming it into a valuable, member-driven resource from day one.

Long-Term Vision: An Infinite Career Compass

The ultimate vision is for this tool to become an indispensable, self-sustaining resource for the entire student community, not just at ASU, but the world. Clubs, student chapters, colleges, full time employees looking for a career shift, everyone can use this tool. As the user base expands and the dataset grows, the AI-powered insights will become exponentially more powerful. Future plans, which will now be driven by direct user feedback from our live ACM deployment, include adding advanced filtering options and features to export personalized career advice.

Base44 Impact Track Details :

The Problem We Are Solving Students today face immense pressure and uncertainty in their job search, often resorting to "resume roulette"—endlessly tweaking their CVs with no real data to guide them. This creates significant anxiety and inefficiency, hindering their ability to succeed in their academic-to-career journey. Our project directly tackles this critical student pain point by transforming the isolated, guesswork-driven process of resume writing into a collaborative, data-driven experience. We aim to replace career anxiety with career clarity, empowering students to learn what companies are looking for and collaborate smarter by pooling their collective, anonymized knowledge.

Our App Solution and Features Our solution, ACM ResumeAtlas, is a live career intelligence platform built and hosted on the Base44 platform. It allows students to safely and anonymously contribute their resumes, which are then aggregated into a dynamic, interactive bubble chart showing real-world connections between companies, roles, top skills, and projects. The core feature is our AI Insight Engine; with a single click, the platform uses the Google Gemini API to analyze the data for a specific role and provides a concise, personalized action plan, acting as an on-demand career coach. To ensure the platform's integrity and build user trust, it is designed with a robust, privacy-first architecture, featuring a human-in-the-loop admin curation system to guarantee data quality and a "right to be forgotten" for users.

Expected Impact and Scalability The expected impact is immediate: to reduce student anxiety and empower them with the confidence to build resumes that are aligned with real industry expectations. The project is not just a prototype; it is already live and being deployed to the ACM chapter at ASU, turning a hackathon project into a real-world community tool from day one. Our platform is designed for massive scalability using a "flywheel model": the more students contribute, the more valuable the insights become, which in turn attracts more users. This model allows ResumeAtlas to grow from a single student club into a university-wide, and potentially global, resource for students and career-shifters looking to navigate their professional journey with data-driven confidence.

Built With

Share this project:

Updates