This is the github repository for Team 15's proposed solution for the Morgan Stanley Alpharetta 2023 Hackathon, created for the Make-A-Wish Georgia foundation: a GPT-based chatbot and web solutions that streamline communication between Make-A-Wish and families.
- Mohammad Arjamand Ali
- Tri Nguyen
- Souradeep Ghosh
- Dheeraj Kumar
- Sparsh N
- Divyalakshmi Varadha Rajan Prem Sudha
- Creating a real-time communication platform for families across Georgia
- Promoting to families not in the program so every eligible child can make a wish
- Connecting families to support resources in their area
- Current availability of volunteers isn't enough to address the full scope of the task, which is the 900 children in Georgia waiting for their wishes to be fulfilled
A chatbot backed by OpenAI's GPT model, that solves communication issues between Make-A-Wish and families by:
- Making communication instant rather than having to wait for a response
- Volunteers' responsibilites become streamlined, and their conributions can now be focused on more targeted and impactful ways
Web solutions to specific challenges faced by Make-A-Wish:
- A document portal allowing families to upoad documents safely and at their own convenience
- A dicussion board for alumni families to connect with each other
A chatbot which utilizes OpenAI's API to fine-tune a GPT model to suit Make-A-Wish and the families' specific needs. WishConnect is capable of conversations and answering questions about Make-A-Wish.
WishDocu is a secure portal though which families can upload their documents with ease. Discussions is a place where families can connect with other families to help them during this hard time.
A Full-Stack Ruby on Rails App that implements WishConnect, WishDocu & Discussions. The Rails framework ensures data security and also provides ActiveRecord, a powerful API for databases.
WishConnect can be rapidly improved with more data and allowing learning during runtime. Features like Authentication, Payment Processing & Mailers can be easily implemented in Rails using Gems.