User Interaction mode
user Interaction mode response
Synthetic Test mode response
Initial prototype

Streamcare: AI + Data Personalized Care

Inspiration

Inspired by the need for precision in patient care, Streamcare was conceived to bring the power of AI to the fingertips of healthcare professionals.

What it does

Streamcare is an intuitive app that allows medical staff to input patient data and receive AI-generated suggestions for medication adjustments, tailored to each patient's unique medical profile.

How we built it

We built Streamcare using Streamlit for a seamless user experience, integrated with the replicate Python package to access the powerful snowflake/snowflake-arctic-instruct AI model.

We had issues integrating streamlit, so we started prototyping with just the vscode terminal
We gradually introduced streamlit and compartmentalised our logic into user experience modes and python modules
The main_app.py streamlit app calls each module depending on which user mode is selected in the sidebar.

Challenges we ran into

One of our main challenges was ensuring the AI model's suggestions were accurate and clinically relevant, requiring extensive testing and iteration.
we also encountered initial difficulties with integrating streamlit with our application logic and Snowflake/Replicate API response

Accomplishments that we're proud of

We're proud of creating an interface that is not only user-friendly but also robust enough to handle complex medical data and provide valuable insights.

What we learned

This project deepened our understanding of AI's potential in healthcare and the importance of user-centered design in medical applications.

What's next for Streamcare

Looking ahead, we plan to expand Streamcare's capabilities by:

updating the training simulation mode,
integrating data retrieval methods with our AI response for more accurate results, and
collaborating with medical institutions for real-world validation.