Goobl

Inspiration

We started this project from something personal. During development, both of us got sick and lost our voices at different points. Noticing changes in our voices led us to research and discover that Parkinson’s disease affects the voice early in the disease. Up to 90 percent of people with Parkinson’s show voice changes before motor symptoms appear, yet most diagnoses still happen after visible symptoms. That gap inspired us to see if voice could be used to increase early awareness.

What it does

Goobl is a web-based tool that records short guided speech prompts and analyzes features like pitch, vocal stability, jitter, shimmer, and rhythm. The system identifies patterns associated with Parkinsonian speech to show how voice changes can reflect early neurological signs. Goobl does not give a medical diagnosis but demonstrates how these patterns could encourage earlier screening and awareness. Users can interact through playful prompts and mini-games that make recording natural and easy.

How we built it

We built Goobl as a full-stack web application. The frontend handles voice recording and interactive prompts, while the backend extracts acoustic features and passes them to a machine learning classifier trained on Parkinson’s voice datasets. Features like jitter, shimmer, and fundamental frequency variation are analyzed to detect irregularities. The app is designed to be simple and accessible, with the focus on engagement and awareness.

Challenges we ran into

Our biggest challenge was frontend development. For the first time, we designed and integrated a full character into our interface, and building around that character was much harder than expected. We struggled with layout consistency, visual hierarchy, and making the UI feel cohesive rather than cluttered.

Accomplishments that we're proud of

We are proud to have built a project that raises awareness of how health conditions can be detected through voice and demonstrates how subtle changes in speech could signal early signs of illness.

What we learned

We learned that voice contains measurable health information and that designing a health-focused tool requires careful attention to user experience. We gained hands-on experience with audio feature extraction, machine learning, and frontend design.

What's next for Goobl

Next, we want to expand Goobl to detect other illnesses beyond Parkinson’s. We also want to refine the UI and make the character interaction smoother and more engaging. Our goal is to continue focusing on awareness and usability while making the app feel approachable and fun.

Built With

• android
• expo.io
• gemini
• google-maps
• ios
• react-native
• rest-api
• supabase
• vultr