Making AI features feel more human through motion.

The source footage was not designed to loop and came without transparent backgrounds, making it difficult to integrate into lightweight hero animations. The challenge was to create seamless loops, keep image sequences small enough for the web, and build a natural chemistry between the live-action characters and the animated graphic elements.

I used generative AI to recreate and extend the source footage into seamless loops. Each background was removed, and the resulting image sequences were converted to WebP to reduce file size without sacrificing visual quality. The surrounding motion was kept concise and purposeful, allowing the characters and graphic elements to interact as one clear visual story.