Maya writes like you've been writing for a decade.
She studies your voice, then ships in it. The team can't tell.
Four things, done relentlessly.
Voice match
Reads 200+ of your posts, decks, transcripts. Locks a voice score. Won't ship under it.
Hook discipline
12 words max. No question openers when you don't use them. No emoji unless you do.
Visual briefs
Every post comes with a card, ratio, and design rationale. Not a wall of text.
30-day calendar
From a one-line prompt to a month of hooks, captions, hashtags and visuals.
While you're asleep, the work is already happening.
Outputs you'll see in your workspace.
- ▸ 30-day content calendar per channel
- ▸ Hooks, captions, hashtags, ALT text
- ▸ Visual briefs (ratios, palettes, type direction)
- ▸ Voice QA on every other agent's output
Maya's stack.
How Maya talks to the rest of the studio.
Hook B locked. Voice score 0.97. 8 words. No emoji. Card is brutalist mono, no gradient.
Caption queued for IG carousel — pls QA the second slide, the word 'leverage' isn't in our last 60 posts.
“If it sounds like AI, you didn't read it before you posted it.”
“Voice is a constraint, not a vibe.”
“Write the hook last. Earn it.”
A few things people ask.
Will it sound like ChatGPT?
Only if you want it to. Most teams can't tell within two weeks of voice training.
What if my voice is unusual?
Better. The more distinctive your input, the cleaner the voice lock.
Can Maya write long-form?
Threads, essays, newsletters, decks. The voice holds.
Posts that don't trigger the 'this is AI' reflex.
Hire Maya and the other seven. Up in an afternoon.
Start a workspace→