Inspiration
Emergencies can happen anytime, and quick access to accurate first-aid information can save lives. We wanted to build an AI-powered assistant that provides instant medical guidance, reducing panic and response time.
What it does
AidWise analyzes symptoms, suggests immediate first-aid actions, provides what NOT to do, and fetches relevant instructional videos. It also displays country-specific emergency numbers for quick response.
How we built it
We used Streamlit for the UI, Mistral-7B-Instruct for AI-driven analysis, YouTube API for video recommendations, and a country-based emergency database for localized support. The backend integrates APIs efficiently to ensure fast response times.
Challenges we ran into
- Fine-tuning AI responses to ensure accuracy
- Improving video recommendations for specific conditions
- Handling different country-based emergency services dynamically
Accomplishments that we're proud of
- Seamless AI-driven first-aid guidance
- Personalized emergency contact retrieval
- A user-friendly and intuitive interface
What we learned
- Enhancing AI prompt engineering for accuracy
- Integrating APIs effectively for real-world applications
- The importance of UX/UI in emergency applications
What's next for AidWise
- Multi-language support for wider accessibility
- Voice-based first-aid assistance
- Offline mode for use in remote areas
Log in or sign up for Devpost to join the conversation.