Alfredo -An AI Nutritionist

Dashboard

Inspiration

About the Project

Inspiration

As a student living far from my family and familiar surroundings, I often found myself struggling to keep up with healthy eating habits. Managing groceries, planning meals, and tracking nutrition became extra challenging without the support systems I had at home. I realized that many other students and young professionals face similar difficulties, especially when adjusting to independent living. This inspired me to create Alfredo—an AI-powered nutrition assistant designed to make food management easy, intuitive, and accessible for anyone, anywhere.

What I Learned

Building Alfredo taught me a lot about user-centered design, especially for voice-first interfaces and progressive web apps. I deepened my knowledge of React and TypeScript, explored real-time data management with Supabase, and leveraged advanced AI capabilities through the Gemini API. I also learned the importance of clear documentation and robust feature planning to ensure a smooth user experience.

How I Built It

Alfredo is built as a Progressive Web App (PWA) to deliver a native-like experience across devices, with offline capabilities for uninterrupted use. The frontend uses React 18 and TypeScript, styled with Tailwind CSS for a modern neomorphic look. The backend relies on Supabase for authentication, real-time data, and storage. For intelligent features, I integrated the Google Gemini API for natural language processing and recipe generation, and the Web Speech API for seamless voice interaction.

Challenges Faced

One of the biggest challenges was ensuring reliable voice recognition and synthesis, especially in noisy environments. Balancing real-time inventory updates with performance was also tricky, as was designing an intuitive interface that works well for both touch and voice inputs. Additionally, making sure the app remained functional offline required careful planning and testing.