CarePlusPlus

Inspiration

Health literacy and mental health problems appear to be increasing in importance in Africa and here in the US. Each year, millions of people, especially young adults in the ages of 18 - 25 are affected by mental health problems and may not have access to high quality health literacy information provided by organizations. Both health literacy and mental health can be very complex. CarePlusPlus, a GPT-powered healthcare super application can simplify the problem. It is affordable, easy to use and provides a toolbox of applications in one place for healthcare interactions.

What it does

CarePlusPlus is a people-centered healthcare SuperApp toolbox for health literacy and education and promoting relationship between physical health, nutrition and mental well-being for optimal overall health.

We are currently starting with two apps in CarePlusPlus SuperApp, namely CarePlus and Moodl apps.

With CarePlus, people will be able to easily access health information provided by organizations.

With Moodl, track your mood and stress level manually and automatically using wearables, IoT with AI analytics, discover mood and stress relief activity recommendations, including connecting with experts or therapists in the field.

How we built it

Health Literacy and Education:

A knowledge base for health literacy and education is setup by uploading documents as resources for the system. Users can interact with this system via a GPT-powered chatbot.

Amazon S3 is used for for storing documents e.g PDF.

Pinecone database is used for storing vector embeddings of the documents uploaded as knowledge base.

Neon, a serverless postgres database is used for storing messages and chats.

Clerk is used for Authentication.

GPT-4o-mini and text-embedding-small models are used for chats and embeddings.

Moodl

Leaflet is used for rendering locations of therapists on a map.

Challenges we ran into

I had a rate limiting issue when using the Open AI text embedding model to store embeddings of the chunked PDFs in the Pinecone database. As mentioned in the error log, I had to apply my debit card to Open AI in addition to the amount provided by the sponsor organization for this challenge.

Accomplishments that we're proud of

What we learned

I learnt the use of the OpenAI SDKs as well as some other databases and frameworks used to build this project.

What's next for CarePlusPlus

A companion mobile application will be released for Android and iOS apps in the future.

Built With

• gpt-4o-mini
• neon
• next
• pinecone
• postgresql
• react
• text-embedding-small
• typescript