CareConnect

Homepage - CareConnect
Dashboard - CareConnnect
Health Vitals IOT - CareConnect
Reports page - CareConnect
AI Health Assistant with RAG - CareConnect
Appointments page - CareConnect
IOT Video Call for CareTeam using Zoom - CareConnect
Patient Information in Settings - CareConnect

Inspiration

Many health applications inspired this project, with features including 5G Connection for Communication between the Hospital's Care Team and patients, and an AI Health Assistant that provides you with information on your health, blood status.

What it does

1/ AI Health Assistant (Powered by Claude 3.5 Haiku with Bedrock - Pinecone Knowledge Base): This feature allows users to get more information about their health status, and provides the latest information based on the user's report trained in the knowledge base.

2/ CareConnect Care Team (Powered by Twilio and Zoom Video Call): This feature allows users to connect with their local healthcare providers and ensure stable connections anywhere in the US. 5G Implementation is in future development.

3/ Appointment Scheduler (Developed with Firebase API): You can schedule your visit with any doctor that fits your schedule and integrates into the main service. Not worrying about missing any doctor visits.

4/ Reports: This is where all the patient documents( health reports, blood test results,...) were hosted, allowing users to access their reports and understand their health condition.

5/ Health Vitals (IOT in future development): You will be able to connect to your wearables and/or phone's sensor to scan your health vitals and detect any health anomaly.

How we built it

We built this project with our front-end in Next.js and React with TypeScript, and the website was hosted with Vercel. Our backend was built mostly with Python (with Flask) and Node.js.

Our project also utilizes the use of Claude-3.5-Haiku in Bedrock Marketplace, utilizing Knowledge Base from Pinecone and Bedrock Agent for our AI Health Assistant in AI Messages. The Knowledge Base initiates the Pinecone Vector Indexes to search the health documents being stored in S3, giving accurate information on the patient.

We also built with different cloud services, including Twilio for Care Team's Messaging and Video Calling Feature, and Firebase App for Appointment Schedules and Reports.

Challenges we ran into

Our initial idea for this project was mainly focused on building healthcare apps that would help people in different countries around the world. This idea was good, but we later changed our focus to AI and 5G, which continues our purpose for this project. We have gone through some challenges in maintaining a good project structure, working with different APIs to build the backend features of the project, and using technologies that we didn't think would be possible (Pinecone for Vector Search, Bedrock Knowledge...).

Accomplishments that we're proud of

We're proud to get the features worked on as we keep building this project. As everything comes together, I finally understood what we've built and the positive impact it could bring to many people in our local community.

What we learned

We learned to manage our project structures and fully built a web application that focuses on communications and health assistance, which helps people understand their health better and supports them wherever they need. We also learned to evaluate our ideas and work with each other.