Inspiration
Drunk driving has been and is still a big problem around the world. After an event involving drinking, people confidently get in their cars confident in their ability to make their way back home. This confidence, however, is yet another effect of being impaired.
What it does
BuzzBuddy measures cognitive ability and motor skills relative to sober baselines to determine whether the user is likely impaired at a given moment. It does this through a proactive AI agent which uses knowledge about a user's physical stature combined with past baseline tests to reason and suggest tasks to figure out whether a user is impaired.
How we built it
We split the work into a frontend and a backend. The frontend uses Swift, an iOS native development language. This was done solely on XCode. The backend uses Python, and calls Digital Ocean's AI tools for our agentic framework.
Challenges we ran into
Nearly every problem we ran into involved Git. Learning how to properly source-control and collaborate with Git was a major part of this project.
Accomplishments that we're proud of
The main accomplishment was getting the AI agent to proactively assign impairment tests and reason through the results. We found that the tests we
What we learned
We learned how to collaborate better and use git properly to work as a team both efficiently and effectively. We learned more about the iOS environment, as well as how to use XCode and develop on our iPhones.
What's next for BuzzBuddy
BuzzBuddy will continue to be iterated upon. The goal is to take this project to a startup accelerator such as Y combinator in the future.
Log in or sign up for Devpost to join the conversation.