Inspiration
CareTV AI addresses accessibility gaps in home healthcare by bringing structured, AI-assisted health and care guidance directly to living rooms via Amazon Fire TV.
What it does
- Designed specifically for Leanback 10-foot Smart TV UI using D-Pad remote navigation.
- Powered by AWS Bedrock (Claude 3 Haiku) for instant, low-latency contextual guidance.
- Features multimodal categories: Household Health, Elderly Care, Livestock Care, and Emergency First Aid.
- Provides dynamic QR codes for mobile connectivity and automated Text-to-Speech (TTS) response playback.
- Includes reliable offline fallback responses and strict non-diagnostic health disclaimers.
How we built it
- Built using Flutter with custom FocusNode TV navigation widgets.
- Connected to AWS Bedrock runtime via AWS Lambda & API Gateway proxy for secure keyless client interaction.
- Designed with high-contrast UI components dynamically scaled for 720p, 1080p, and 4K TV displays.
Challenges we ran into
- Handling 10-foot UI remote focus highlights cleanly across various screen densities.
- Ensuring compliance with store and hackathon health guidance policies by integrating clear disclaimer-based UX.
What's next for CareTV AI
- Adding multi-user household profiles and expanded voice interaction capabilities via AWS Bedrock Agent endpoints.
Built With
- amazon-web-services
- android
- aws-bedrock
- dart
- fire-tv
- flutter
Log in or sign up for Devpost to join the conversation.