For my work at Hugging Face I need to create new generative audio/image/video demos, but this is not efficient as it requires building stuff from scratch using Gradio and/or React, re-inventing the AI wheel and pipelines all the time. Instead it is more efficient if I use Clapper as a generative AI video framework, as it already contains 90% of what I need.