Find the gold in your long-form content.
Extract viral moments from long podcasts.
Podcasters record 1-2 hour episodes but struggle to promote them. The best content is buried in hour-long recordings. Manually scrubbing through audio to find clip-worthy moments takes forever. Most podcasters never repurpose their content, missing huge audience growth opportunities.
Upload a podcast episode, AI transcribes it, identifies the most engaging 30-90 second segments, and generates shareable video clips with auto-generated captions. Export ready for TikTok, Instagram Reels, and YouTube Shorts.
Upload episode
Audio or video file
AI finds highlights
Best moments detected
Export clips
With captions, ready to post
Create a Next.js app for podcast clip generation. Features: - Dashboard: uploaded episodes, generated clips - Upload: audio/video files up to 2GB - Processing status: transcription progress, clip detection status - Clip editor: trim, adjust, preview - Export: download clips in multiple formats Use Supabase for auth. Mux or Cloudflare for video processing.
Build the AI highlight finder: 1. Transcribe audio using Whisper API (with timestamps) 2. Send transcript to Claude with prompt: "Find the 5 most engaging, shareable moments in this podcast. Look for: surprising statements, emotional moments, actionable advice, controversial takes, funny exchanges. Return timestamp ranges and why each is clip-worthy." 3. Extract those segments from original audio/video 4. Rank by predicted engagement Let users mark favorites and reject unwanted suggestions.
Create the video clip with captions: 1. Extract clip segment from source 2. Generate word-level captions from transcript 3. Apply caption style: animated word highlighting, emoji reactions 4. Add branding: logo watermark, podcast name 5. Format for platform: 9:16 for TikTok/Reels, 1:1 for Instagram 6. Render using FFmpeg or Remotion Templates: choose from pre-designed caption styles and layouts.