Finacial Fables (Group 20)

Choose financial topics, our AI creates stories and images, enhancing kids' financial understanding in an enjoyable learning platform

Comment

Inspiration

Do you ever think about how easier life would be if only you'd been taught about finances while you were younger? Financial literacy is a vital life skill often overlooked in traditional education. We were inspired to bridge this gap and make learning about money engaging and fun for children. The idea of Financial Fables was born from the desire to empower the young generation with practical financial knowledge, setting them up for a brighter future.

What it does

  1. The home page basically informs the users about the functioning of the app. It also has a list of popular topics that users might be interested in learning.
  2. The next page, new Story, takes inputs from the users(children or their parents) - Their age (so the content would be curated according to their age), the character's name, and description (this character would be the protagonist in the story).
  3. From the inputs provided, Gpt4 writes a story prompt and divides it into 10 frames for the next step. Next, the 10 prompts are sent to Dalle to generate appealing and interesting images for the story prompt.
  4. The user can view the story and learn the basics of finance from day-to-day life situations.
  5. The users can also see the stories generated by them and other users in the My stories and explore and learn as they go!

How we built it

  1. Conceptualization: We started by defining the project's scope, identifying our target audience (children aged 6-14), and outlining the key features.
  2. User-Centered Design: We focused on creating a user-friendly interface that would be easy to use. This involved a simple character creation process, where kids could create their characters through text input and a personalized story request system.
  3. We started by using OpenAI and Langchain to get a large language model to produce a 10-frame story. For each new frame (section of the story) we prompted the model to use #NEWFRAME so that we had a repeated string. We used this to split the generated text into chunks, one for each frame of the story.
  4. For each frame of the story, we prompted the large language model to generate a description of an image that could accompany each frame. We then sent this image description as a prompt to OpenAI's Dalle-2 to generate a still for that section of the story.
  5. At this point, we were able to generate a 10 frame story with 10 images, one to accompany each frame. We then decided that to add a database that stores the names of the stories and the text associated with each individual frame. We also had to upload the generated images to the database so that they could be quickly recalled for future review.
  6. Finally, we created a 'my stories' page, which allowed the user to read and look at their generated stories for later use/ read.

Challenges we ran into

Since we four are not from a tech background and it's our first time participating in a hackathon, it was challenging to finish our project from scratch. Some of the problems we faced are-

  1. We needed to teach ourselves how to use a suitable database for saving the stories generated by users so that we could display them for others to view.
  2. OpenAI returns images in a new datatype we didn't know how to work with.

Accomplishments that we're proud of

  1. This is the first time we have built something! We're proud that we were able to identify a gap and propose a working solution for it.
  2. Teamwork! We four barely know each other so it was great to come together as a group and contribute, each in our own way!

What we learned

  1. We learned a lot of coding as we were building the web app.
  2. How to connect databases to our app to save the generated content.
  3. How to choose a topic that matches the skillsets of all the group members, so that it is easier

What's next for Finacial Fables

  1. We want to improve the UX of the application so it's more interactive and engaging for children.
  2. Add a log-in page so we can track our customers and let them save their stories as they learn.
  3. Provide customized resources for every topic the child wants to learn about so that they can move ahead with their learning journey.

Built With

  • dalle
  • langchain
  • openai
  • python
  • streamlit
  • supabase
Share this project:

Updates