@sh_reya
we've been iterating a lot on how DocWriter internally represents a user's writing style! the naive approach is to dump all a user's prior writing in context and pray the AI figures out how to sound like them. this doesn't really work. and the user can't just manually specify their style either, because nobody knows what their meaningful writing preferences (for an AI agent) actually are. so we've built a multi-agent pipeline that runs lexical, grammatical, and discourse analyses on the user's writing, extracts specific patterns with evidence, and then surfaces them in a UI where the user can confirm or reject each pattern by comparing two passages side by side TBD on how well it performs in our next round of user studies 😆 it feels like we are really trying to do anything possible to avoid fine-tuning, but perhaps we need to bite the bullet eventually