🛠️ How We Built It
We built ElderEase as a full-stack AI-powered web app designed to simplify healthcare access for senior citizens. Here's the tech stack that made it possible:
⚙️ Backend:
- FastAPI to build a scalable, asynchronous REST API.
- Google Gemini Pro & Gemini Vision for all AI-driven features like document summarization, natural conversations, journal title generation, and emotional tone detection.
- FastAPI to build a scalable, asynchronous REST API.
🎨 Frontend:
- Next.js for server-side rendering and routing.
- Tailwind CSS and Framer Motion to create a soft, elegant, and animated user experience that's friendly for seniors.
- SpeechRecognition API and Text-to-Speech for voice input and audio playback.
- Push Notifications to remind users of prescription times.
- Next.js for server-side rendering and routing.
☁️ Backend as a Service:
- AppWrite was used to manage file storage, database, and authentication in a streamlined way.
🧱 Challenges We Ran Into
- 🔁 Gemini model versioning: Keeping track of different Gemini API versions (Pro vs Vision) and the required prompt formats took some trial and error.
- 🧓 Designing for elderly users: We had to think deeply about spacing, font sizes, icons, and accessibility, making sure the experience was warm and not overwhelming.
- ⚒️ Backend reusability: We aimed to keep our backend generic and stateless so it can power future mobile apps or even smart kiosks in clinics or pharmacies.
📚 What We Learned
- 🤖 How to work with Google Gemini APIs for text and image understanding at scale.
- 💬 Building interactive chat UIs that retain conversation context and provide helpful follow-ups.
- 🔄 Managing state, routing, and file handling in a full-stack production-grade app.
- 🧏♂️ The importance of simplicity, empathy, and accessibility when building for a senior audience.
Log in or sign up for Devpost to join the conversation.