Uplift

Inspiration

A new viral trend on social media short videos is recordings of text messages aided by an app to "pick someone up" or flirt with them. Our group found this concept to be interesting, but more useful and applicable in a different context: helping our friends when they are going through a challenging time. We all have faced situations where we had to be there for a friend but were unsure about what exactly to say. How can we support our friends no matter what? This inspired us to create an app that helps with exactly that.

What it does

With Uplift, you can take advantage of AI to help craft the best response to a friend to be there for them in the best way possible. We have a simple easy-to-use user interface through which the user can provide our app a little more context into what the friend is facing, how the user wants to address the issue, and more information. We feed this data into AI and provide a list of sample responses that the user can respond with or take inspiration from, which can be copied directly from our app.

How we built it

We used Penpot to build the wireframe, React Native to build the UI, and LangChain API and Open AI API to process the user situations and return sample responses. We also tried to implement Tesseract JS for optical character recognition from an image.

Challenges we ran into

Two of our group members had to learn React the day we started coding. In addition, we had to find a way to access the LangChain API because of OpenAI's token limit. Learning Tesseract JS on the job was also a skill necessary to create this app.

Accomplishments that we're proud of

We were able to learn the basics of React Native, Penpot, and Tesseract JS in a short timeframe, and create a beautiful-looking app with most of the functionality we were hoping for from an MVP.

What we learned

We learned more about collaborating on code with each other, how to standardize coding on a large project like this, the importance of listening to each other's ideas and asking for feedback and advice, and new technical skills to use in the future.

What's next for Uplift

We plan to implement uploading screenshots, which we do have the API for but need to work on fully integrating with the app. We also hope to create an interface to give feedback, train our AI, and increase the user-friendliness of our interface by adding more mobility.

In exploring mental health resources, we found most apps require payment in return for professional services or untrained user volunteers, which is not an entirely viable solution. Therefore, we want to fill this niche in mental health services on a larger scale by utilizing the expert element that AI brings along with the human perspective from users and volunteers who personalize conversation.

[link]https://www.youtube.com/watch?v=76T-BBqLBeY --> alternative link if video submission does not work

Built With

• figma
• langchain
• openai
• penpot
• react-native
• tesseract.js