Companion

• 

  Figma representation of Companion

Inspiration

Companion The goal for this project was a platform that provided a safe space for people to reach out to and interact with while have a means of getting immediate attention.

What it does

Our idea behind this was a Web App that would serve as a companion to patients, and people in general who need someone to talk to. The chat bot can help maintain basic conversations and the platform will provide a way for people to contact a nurse immediately if they have an emergency, or need medications. Lastly, as a means of providing first responders with a greater sense of reactive ability, we hoped to incorporate a sentiment analysis of the User input as a means of giving more insight into what the person might be feeling.

How we built it

Currently, we have a working chatbot that we build using Google's PaLM API, along with a hugging face sentiment analysis built using the j-hartmann emotion-english-distilroberta-base model and over 20,000 observations. We built both of these using the platforms on Google's Makersuite and HuggingFace spaces respectively.

Challenges we ran into

In the process of building a chatbot and doing sentiment analysis of the input text from the users, we ran into challenges while integrating the two tools together while further embedding them into a webpage.

Accomplishments that we're proud of

We're proud of being able to utilize the several tools that were taught during the course of the hackathon including Google's Makersuite and HuggingFace APIs. We're proud of being able to create a successful chatbot and sentiment analysis!

What we learned

We learned how to make use of several tools such as huggingface as well as Google's Makersuite. Learning how to use them for the first time and having them work was motivating!

What's next for Companion

We hope to incorporate the chat bot we built into a Web Application that would provide users with an interactive interface where they can communicate with the chatBot or a nurse. We also hope to combine the sentiment analysis we did onto the backend of the Web App.

Built With

• huggingface
• makersuite
• palm
• python