Studio automation, video processing pipelines, and AI scripting
Python handles the studio's automation layer — the scripts that run in the background and make repetitive work disappear. higgsfield-compress.py processes raw video exports from Higgsfield into web-ready files: right codec, right bitrate, right container, no manual steps. A batch job that used to take an hour runs unattended in minutes.
Whisper handles transcription for video content — lecture recordings, client interviews, voiceover drafts. The pipeline runs locally: audio in, clean transcript out, ready for editing or embedding. mac_writing_assistant.py extends this into a dictation workflow, letting long-form writing happen by voice and land as structured text.
Python also powers the glue code between services — Notion API calls, file watchers, scheduled reports. It's not the language of choice because it's elegant; it's the language of choice because the libraries for AI, media processing, and API work are all there, and Claude Code can write and iterate on scripts quickly when the need arises.