Case study / 03
An AI podcast studio: write a script to an exact length, voice it with AI or your own voice, mix it on a real timeline, master it to broadcast loudness and publish the feed.
A podcast has a slot. The script should fit it on the first try.
Scripts are written to the target runtime and held within ±5% by extend-and-trim passes, so a 20-minute slot gets a 20-minute episode.
Speakers, delivery and sponsor breaks are set beside the script, with a segment inspector for any single line.
AI output is a first take. Creators need to be able to fix it.
Voices, music, effects and ambience sit on their own lanes with volume, pan, envelopes and automatic ducking under speech.
Every clip can be trimmed, moved or regenerated without starting over, and the mix renders on the server.
Trust comes from showing the work, not claiming it.
Mastering lands at −16 LUFS with peaks under −1 dBTP, and the report says what that means in plain words before showing the numbers.
A before-and-after excerpt and waveform let creators hear the difference, and WAV, MP3 and AAC export from the same screen.
CastGenX is live at castgenx.com. Next is the public launch and a lifetime deal on AppSumo.
