Automating Content Repurposing with n8n

Hey WebNutch community, I wanted to share my experience with building an automated content repurposing pipeline using n8n. After months of iteration, I've finally got a workflow that connects Google Drive, Whisper, GPT-4, and FFmpeg to convert long videos into 10 short clips ready for TikTok, Reels, and Shorts. One of the challenges I faced was dealing with the classic problem of 'decapitated' faces when converting from horizontal to vertical, but I was able to solve it by including face detection for 9:16 cropping. I also encountered an issue with the FFmpeg local server cache reusing old horizontal videos when passing vertical=true, but I was able to resolve it by adding dimension validation with ffprobe before returning from cache. If you're interested, I've shared the JSON file for import on my GitHub page, and I'd love to hear any feedback or suggestions on how to optimize the pipeline. I'm also looking to promote this workflow to potential clients on my website, where I've included an installation manual and the option to schedule an installation with me. Check it out and let me know what you think!