video1.0<iframe src="https://www.loom.com/embed/c44968dd41aa48eea83301959950262e" frameborder="0" width="2214" height="1660" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>16602214Loomhttps://www.loom.com16602214https://cdn.loom.com/sessions/thumbnails/c44968dd41aa48eea83301959950262e-fe352dba9dd79cab.gif104.967This Loom explains how Sissy uses Revert to draft messages in her own voice instead of a generic tone. She trains a small base model (Quen 3.8, about 7) using thousands of threads from Cloud Code sessions plus Messages, X, and Gmail, so outputs sound like her. She describes a workflow where her model and a friend (Yana) can respond quickly to suggested messages and draft emails in Gmail and other places, including rephrasing highlighted text in her tone.