Lazy Care

LazyCare is an AI-powered health assistant that provides personalized lifestyle recommendations through text-based interactions.

Comment

Inspiration

Our inspiration for LazyCare stems from a reality we all face: while health is universally important, quality healthcare guidance may be inaccessible for many. Appointment delays, high consultation costs, and overwhelming generic online information can make timely, personalized care difficult.

What it does

LazyCare is an AI-powered health assistant that provides personalized lifestyle recommendations through text-based interactions. The app enables users to manage their profiles, engage in AI-driven health conversations, and maintain a chat history. A custom-trained model was also developed but not yet integrated.

How we built it

We build with Next.js for the frontend and Node.js (Express) / Python (FastAPI) for the backend, it leverages tiny LLaMA to analyze user-inputted health data, including sleep patterns, weight, and overall well-being.

Challenges we ran into

We need to train our Ollama AI from scratch to interpret symptoms and provide medically sound, personalized advice rather than generic recommendations.

AI lacks memory, we must implement a custom memory system to track user health history, ensuring the AI continuously tracks the user's progress and care for the user.

What we learned

We gained valuable experience in AI model pretraining and fine-tuning for healthcare applications. and we also learned how to build a complete project from scratch, managing the entire development lifecycle from concept to deployment.

What's next for Lazy Care

In the future, Lazy Care can implement voice input and speech-to-text functionality to make the app more accessible and convenient. We can also develop real-time data integration with popular health-tracking devices and apps and add more sophisticated health analytics to identify trends and provide more proactive recommendations.

Built With

Share this project:

Updates