Podcast Summarizer

National sports podcasts mention your team for maybe a minute buried in a 90-minute episode. This pulls out just that part.

Takes any podcast URL, transcribes the full episode with Groq Whisper, and summarizes just the content relevant to a specific team using Llama 3.3 70B. The pipeline handles episodes up to 75 minutes by splitting audio into chunks under Groq's upload limit, then running a two-pass summarization for longer transcripts.

Transcripts are cached locally so re-running with a different prompt is nearly free. The whole thing is a single bash script with no package manager — just curl, python3, and ffmpeg. Costs almost nothing to run.