💡 Inspiration
Existing health apps often silo physical and mental well-being, tracking metrics in isolation. We asked: what if we could unify this data, pulling from wearables and daily self-reflection, and use the power of AI to uncover hidden connections within our own health journey? We envisioned a tool that transforms scattered data points into a meaningful personal health narrative.
⚙️ What it Does
HealthSync bridges the gap between quantitative wearable data and qualitative self-assessment. Our Flutter application integrates with Apple Health (via HealthKit on iOS devices) to fetch objective metrics like sleep duration, activity levels, and heart rate. Users supplement this by easily logging their daily mood, energy levels, specific symptoms, and contextual notes.
HealthSync then securely processes this combined data, leveraging Google's Gemini AI to analyze correlations and generate personalized, easy-to-understand health narratives. These insights help users visualize patterns they might otherwise miss – such as how sleep quality impacts their energy the next day, or how specific activities correlate with reported symptoms – empowering them to make more informed lifestyle decisions and facilitating better communication with healthcare providers.
🛠️ How We Built It
We initially planned an iOS-native Swift application but quickly pivoted to Flutter. This allowed our team members, working across both macOS and Windows, to collaborate effectively on a single codebase.
Our core stack includes:
Flutter/Dart: For the cross-platform mobile application frontend. Firebase Authentication: For secure user login (a pivot from our initial Auth0 attempt). MongoDB Atlas: As our flexible NoSQL database for storing user logs and health data. Google Gemini API: The AI engine analyzing data and generating personalized insights. Apple HealthKit Integration: Via Flutter's health packages to access iOS health data.
Challenges We Ran Into
Authentication Hurdles: Our biggest technical challenge was authentication. We dedicated many hours trying to implement Auth0, but encountered persistent redirection issues on iOS simulators/devices that proved insurmountable within the hackathon's tight timeframe. This forced a critical, late-stage pivot to Firebase Authentication, which we successfully integrated. Task Delegation & Blockers: Like many hackathon teams running on caffeine and little sleep, managing task dependencies and keeping everyone productive when inevitable 'blockers' arose required constant communication, patience, and adaptability.
🏆 Accomplishments That We're Proud Of
Despite the challenges, we're incredibly proud of delivering a functional MVP. This is especially significant given that most team members had limited prior experience with Flutter or MongoDB. Our ability to rapidly pivot on a core technology like authentication while maintaining team cohesion and project momentum was a major success. We truly enjoyed the collaborative process, tackling complex problems together and celebrating each milestone as HealthSync took shape.
📚 What We Learned
Gemini is Powerfully Versatile: Beyond being the AI engine for HealthSync's insights, Gemini significantly boosted our development productivity. We extensively used it via tools like the Roo Code extension, feeding it code snippets for debugging, explanation, and generation – it felt like having an AI coding partner. The Power of Connected Data: Even during testing, the app started revealing surprising correlations in our own health data just by combining simple daily logs with passive wearable metrics. It validated our core concept firsthand! Adaptability is Key: Successfully navigating the authentication pivot reinforced the importance of being decisive and flexible when facing unexpected roadblocks under pressure.
🚀 What's Next for HealthSync
This hackathon project is just the beginning! We genuinely believe in HealthSync's potential to empower users by making their health data understandable and actionable. Our immediate next steps involve refining the user interface, potentially expanding data sources (e.g., adding Google Fit support for Android users via Flutter), and further enhancing the AI's pattern detection capabilities. We're excited to keep grinding and bring our vision to fruition!
Built With
dart firebase flutter gemini github google-cloud healthkit mongodb-atlas
Log in or sign up for Devpost to join the conversation.