TAMU Banners

This program uses machine learning so that you can get plugged in to an A&M community easier than ever!

Comment

Inspiration

As brand new Aggies, we didn't know where to find "our people" for the first couple of weeks. The process of integrating into the A&M community is an exciting prospect, but under all the optimism, sometimes it was exhausting and discouraging, especially when we didn't know where to look.

What it does

TAMU Banners helps students discover organizations that match their interests. Users enter keywords, and the app displays personalized cards from over 1,300 campus organizations. As users like and dislike their feed, the machine learning model continuously refines its recommendations, generating more personalized responses.

How we built it

React for front end, FastAPI for back end. Google, Microsoft, and OpenAI for content generation and troubleshooting. The machine learning algorithm uses logistic regression models and cosine similarity formulas to determine scores. Python for web scraper.

Challenges we ran into

Learning how to host a website for the first time was quite difficult, as well as integrating the different components of the full-stack application. We had started with no knowledge of full-stack web development, server hosting, and machine learning models, but quickly learn to integrate those components into our product.

Accomplishments that we're proud of

This is our first Hackathon, and our team caught on fast to important concepts despite technical challenges. We built a fully functional machine learning algorithm and a thorough web scraper that returned data from 1300+ organizations' webpages.

What we learned

We learned about the process of full-stack development and prompt engineering, as well as tools like Git, React, Streamlit, and FastAPI.

What's next for TAMU Banners

We aim to further our mission of helping freshmen find their community. Next steps include creating user profiles, friend integrations (see which orgs your friends like), and organization dashboards to upload promotional videos, application deadlines, and exclusive events (who doesn’t love free food?). We'd make it effortless for users to connect with the community and for the community to connect people. Long term, we plan to build a mobile app with user authentication and a sleek, Tinder-style swipe interface for exploring organizations effortlessly.

Built With

Share this project:

Updates