Mood Journaling

Record your experiences in an AI-powered scrapbook-style journal.

Comment

Inspiration

We were inspired by the power of combining technology and emotional expression. We wanted to create a platform that not only allowed users to journal their feelings but also provided them with a visual representation of their emotions. This project bridges the gap between art and technology, offering a unique and personalized experience for users to reflect on their daily moods.

What it does

Users can input a short entry about their day, which is stored in the user's journal along with a corresponding AI-generated image. Users can then look through their journals to see entries from previous days or check the calendar to get a broad overview. The app aims to create a messy, scrapbook-like experience, adding a touch of artistry to the process of emotional expression.

How we built it

Front End

  • Reactjs
  • CSS
  • Javascript
  • Framer-Motion

Back End

  • Flask
  • Python
  • PostgreSQL
  • Digital Ocean

APIs

  • openAI
  • stability.ai

The front end was constructed with react, vanilla CSS for styling, and framer-motion for animation. As for the backend, entries are fed to our flask backend through fetch requests. The backend sends user-inputted entries to GPT-3.5, which generates a stable diffusion prompt to match the entry. This prompt is then sent to stability.ai, where an image is generated by Stable Diffusion. The resulting image and original user entry are stored on a PostgreSQL database, to be retrieved at a later time by the user.

Challenges we ran into

We struggled to get our various AIs to cooperate with our requests, as we were passing data first through GPT-3.5 and then to Stable Diffusion. As a result, our results were initially wildly inconsistent, but after many attempts and failures, we managed to get consistent results. Pinpointing the feel of a cozy environment was also difficult, especially with our pseudo-3D design, but thanks to our in-house artist we were able to sufficiently achieve the desired feel.

Accomplishments that we're proud of

We are proud of our AI prompt engineering, which was utilized extensively to generate the desired outputs. We also love the overall feel and look of the app, all thanks to the hand-drawn art from one of our team members.

What we learned

As you can probably guess, we learned a ton about prompt engineering. We can confidently say that we will use the prompting skills we learned during this hackathon in future projects and hackathons. We also learned a lot about working as a team in programming, whether it be problem-solving or resolving merge conflicts. The hackathon was overall an extremely educational experience for all of us

What's next for Mood Journaling

We see a lot of potential in this app, and we hope to deploy it for general use. Of course, we will first touch up the UI and ensure the backend is both scalable and secure, but after that we will see about deploying our app for widespread general use.

Built With

Share this project:

Updates