Inspiration
In the rapidly developing world of sports, data analytics have become a game changer, but here’s the catch: they’re only for teams who can afford them. As high school students, we recognized a massive gap in resources between high-funded and underfunded teams, regardless of high school or college level sports. This is especially striking in underserved communities.
“I truly believe that if you pursue something with all your heart, the universe will align to make it happen.”
It is this disparity that limits opportunities, exposure, and scholarships for talented youth athletes. It is with a smile that we present RFX, our solution to bridge the gap and level the playing field.
What it Does
RFX leverages supervised machine learning models and real-time statistical analysis to generate predictive insights on opponent performance and team strategy optimization. By ingesting historical game data and applying TensorFlow-based models, the platform identifies behavioral patterns, play tendencies, and performance deltas across teams. These insights are surfaced through a user-centric React interface and updated in real time via Firebase integration. Built with accessibility in mind, RFX features a lightweight frontend architecture optimized for low-latency environments and older hardware, ensuring usability across a broad range of devices. The system translates raw, unstructured data into context-aware, actionable recommendations—allowing coaches and athletes to make data-informed decisions, improve pre-game planning, and adapt strategy mid-game, regardless of funding level or technical background.
How We Built it
RFX was developed as a full-stack web application designed for performance, scalability, and accessibility.
On the backend, we used Python to extract, clean, and normalize historical sports performance data from public and semi-structured sources. We applied TensorFlow to train supervised machine learning models that analyze team behavior, player statistics, and game outcomes to predict opponent strategies. To enhance these predictions, we integrated Palantir’s Foundry generative AI API, which provided contextual recommendations and strategic insights using historical and live data inputs.
The frontend was built using React.js, chosen for its modular structure and ability to create a responsive, dynamic user experience. We prioritized usability, designing an interface that’s intuitive for users with limited tech backgrounds—primarily high school coaches and athletes.
We utilized Firebase for real-time database management, authentication, and cloud storage. This allowed for fast syncing of game data, secure access control, and cross-device persistence. Bootstrap was implemented for a responsive and lightweight UI, ensuring compatibility with low-bandwidth networks and older hardware commonly found in under-resourced environments.
Throughout development, we maintained a component-based architecture, used Git for version control, and collaborated through a CI/CD pipeline to enable rapid iteration and testing. Our stack was chosen not just for technical performance, but to ensure the platform remains lightweight, maintainable, and globally deployable.
Challenges We Encountered
Data Access & Reliability: We’ve recognized that many underfunded teams also don’t have access to consistent or structured data. We tackled this by combining public data scraping tools with normalization algorithms to clean and prepare usable datasets.
Design Simplicity: Making advanced analytics approachable to non-technical users required ongoing user testing and design iteration.
Accomplishments We're Proud of
One of our proudest accomplishments was building a fully functional, end-to-end sports analytics platform in a limited time frame—designed specifically for teams with limited access to resources. We successfully integrated Palantir’s generative AI API to deliver real-time strategic insights, a feature typically exclusive to professional-level platforms.
We developed and deployed a clean, intuitive frontend that’s accessible to users with little to no technical background, all while maintaining responsiveness and performance on low-bandwidth and outdated devices. This achievement reflects our commitment to creating inclusive technology, not just powerful software.
We’re also proud of the custom ML pipeline we built using TensorFlow, which allowed us to extract meaningful patterns from incomplete and messy datasets—a technical challenge that required creative problem-solving and collaboration.
Above all, we’re proud that RFX stays true to its mission: delivering cutting-edge tools to underfunded teams, closing the gap between talent and opportunity, and proving that impactful innovation starts with empathy.
What's next for RFX
RFX isn’t just software, it’s a step toward equity in sports. Just as Malala Yousafzai fought for equal access to education, we’re doing our part to ensure young athletes have equal access to opportunity on the field.
Our next steps include expanding RFX to support a larger array of sports beyond our current model and forming partnerships with schools, nonprofits, and community leagues nationwide. We aim to refine our AI models to enhance our AI models by integrating richer, more diverse data sources to improve prediction accuracy and personalization.
Accessibility remains at the core of our mission. We’re actively exploring offline functionality and multilingual support to make RFX accessible in low-connectivity and global contexts. With the right support, we envision RFX expanding globally, leveling the playing field for athletes everywhere, regardless of geography or income.
Log in or sign up for Devpost to join the conversation.