StorySimplify

• 

  main simplify

• 

Inspiration

Inspired by our various learning styles, we decided to create an AI browser automation that allows users to access news articles in whichever learning style they best understand.

What it does

StorySimplify converts news articles into simplified text for any grade level, a children's picture story book, a podcast, and a quiz, for all different learning styles.

How we built it

We built it using Python, JS, HTML, CSS, SQL, SupaBase, Gemini API, and ClipDrop API. We also fine tuned ML models to help classify the reading difficulty of various texts.

Challenges we ran into

We ran into challenges of finding image generation LLMs that didn't have a rate limit but still have high quality image generation while also factoring quick generation times. We also ran into challenges with training the text difficulty classification model.

What's next for StorySimplify

From fine tuning models for this project, we noticed a major gap in datasets for training language density and difficulty classification models. We want to take further steps in researching and developing datasets in this field to help train future ML models.

Built With

• css
• gemini
• html
• javascript
• python
• sql
• supabase