Inspiration

I wanted to visualize a few books i was reading like a timeline - instant movies, now double it with making it interactive and explore the book with questions the pdf didn't even have the knowledge of,

there's more but im in a rush to finish the submission

What it does

It takes an PDF and visualizes them, basically creating instant movies for the pdf, there are subtitles capabilities, background music, generative art and much more

How we built it

created a simple landing page where we can import and select the text on the pdf simply by highlighting it (in future to be converted into a timeline feature) and that selected portion first shifts to the right side so we can look at the text before creating the videos

The video generator is the main ingredient as it has multi-agentic workflow

simply put it's input > Story + Vibe of the story + Voiceover (openai) > Voiceover audio (11labs) > Prompts for each scene based on all the previous data + User INput (openai) > Generating the media (fal ai) > Stitching together + Creating subtitles + sequence setting etcc etc (python, moviepy, ffmpeg etc)

used Lovable for the front end help created variations and picked the best

Challenges we ran into

  • choosing a name
  • literally every step

Accomplishments that we're proud of

  • video generator engine

What we learned

AI is the future and building things

What's next for Kiddle

The core functionality can be expanded laterally across various sectors but first i'd like to crack into the children's audible alternative niche

then maybe expand into research help, real time news, presentations etc

Technically a lot of features need to be implemented mainly the timeline feature and a ton of optimizations as rn its just spaghettiii

PEACE OUT

Built With

Share this project:

Updates