HappyMobile

Inspiration

The challenge of how can T-Mobile know how their customers feel in real time was something we thought was interesting and could be expanded on. Customer satisfaction can be measured through many ways, whether that be through surveys, complaints, or social media posts. But by then, the problem has already spread. What if we build a site that gives T-Mobile a live view on customer happiness, so T-Mobile can act before issues escalate? And with that ideology, HappyMobile was born, a data-driven way to listen, learn, and improve customer experience.

What it does

There are 2 different ways for the site to be used. One through the User, and one through the Admin. Users can login, rate their happiness from a scale of 1-100, report their issues, and explore interactive mapping which shows happiness and problems by region in the United Stats. Admins, while having access to everything the user has access to, can also view and manage every single problem report log sent in by users, create alerts for major updates or outages, and monitor the nationwide happiness index and server status. The app also includes a LLM chatbot, which assists the users with support and service questions related to T-Mobile or Networks in general. Behind the scenes, we used lexicon-based sentiment analysis on Reddit post data and continuously updates the overall Happiness Index based on how customers are feeling online.

How we built it

We used React.js for the fronted, building it page by page, and using libraries such as Framer for animations. We connected it using Node.js and Express.js to our databases which are running through Firebase all to get a working application. All sorts of data, such as happiness index, user authentication, and problem logs are all saved onto our databases, and can be added or edited directly through the frontend. This way all data is saved, and results change in real time with more users on the site.

Challenges we ran into

Many many many errors in both the frontend and the backend. It was a tough challenge to get both the folders to work together, and required A LOT of debugging. Honestly through the early morning, we spent around 2-3 hours just fixing errors until the application finally decided to work. There was also a problem with one of our teammates not having access to a working backend, and required him to think of outside the box solutions to continue getting work done and not waste time.

Accomplishments that we're proud of

It works!!!! We got a polished application running and deployed, with a interactive map feature, animations on the frontend, and a sentiment analysis model which web scrapes reddit data on real time. These features were originally just ideas, but after implementing them, we feel proud in what we did.

What we learned

Since this is all of our first hackathon, it's essentially the first time we worked with databases, frontend code, and integrating AI models. It was a fun experience, and besides the technical skills, we also learned how to essentially stay up for 24 hours and not go crazy while debugging code. A skill all programmers should eventually learn.

What's next for HappyMobile

As of now, we have plans to improve the frontend UI and mapping feature. The map we want to make more complex, and incorporate traveling to major cities rather than just states. And as for the design, we want to make it sleeker and more appealing towards the younger generations. We will definitely continue working on this, making it a top priority for something T-Mobile will be intrigued by.