A complete tour of what Avatorial does — and, just as deliberately, what it refuses to do.
The single source of truth behind every draft. You can read all of it, and edit all of it.
Upload your official LinkedIn data export (the "get a copy of your data" archive) and connect Reddit to import your posting history. No scraping — your archive, your consent.
Your posts are ranked by recency and engagement; you confirm the 3–6 that sound most like you. These rotate through every generation as few-shot examples.
An AI pass distills how you actually write: sentence rhythm, paragraph shape, vocabulary, opening and closing habits, emoji and punctuation, formality, humor, hashtag behavior. Fully human-editable.
Stage one asks your goals, red lines, and tone. Stage two generates ~15 follow-ups that probe tensions between what you said and how you actually write.
A free-text instruction injected verbatim into every generation. "Never say thrilled. Prefer stories over advice." Edit it any time; it wins over everything else.
Your edits and reject reasons accumulate and can be folded back into the style card on demand — the profile gets sharper the more you review.
Set a monthly post budget and a variation mood — Steady (±20%), Varied (±35%), Spontaneous (±60%). Avatorial casts a 4-week calendar: weekday-weighted, inside your active hours, silent (or themed) on your holidays. You approve the plan before a single word is generated. Slots are invitations, not quotas — a weak draft means an empty slot, never filler.
Phone-first: snap a photo, add a one-line hint, done. The hint is the creative brief, the photo is the evidence. A finished draft lands in your review inbox — capture never touches the scheduler.
On Reddit, Avatorial discovers conversations matching your topics and drafts a substantive comment for the best ones. On LinkedIn, paste any post URL to get a reply draft — we never read your feed programmatically. Suggested, never auto-sent.
Between generation and your inbox stands a judge.
Every draft is scored against your own profile on four axes: on-voice, substantive, non-generic, safe. Not a generic quality bar — yours.
A failing draft gets regenerated with the judge's notes, up to three attempts. Still weak? The slot is dropped. An empty slot is a designed outcome; a forced post never is.
When you reject with a reason — wrong tone, too salesy, factually off — that reason drives a targeted regeneration, and feeds the long-term profile.
Company pages carry Drafter, Approver, and Page-admin roles. Any single approver may approve (including their own drafts); only page-admins manage roles and connections.
No role exists that grants anyone power over another person's personal account. That's a database constraint, not a policy promise.
Every approval, publish, role change, and account connection is recorded — who, what, when — and visible to page admins.
Personal LinkedIn publishes via LinkedIn's own "Share on LinkedIn" OAuth. Reddit posts (text and native image posts) via the Reddit API with a disclosed bot identity. Company pages via the Community Management API. Never your session, never cookies.
Approved posts with a future slot publish themselves at slot time — approval is re-verified at the moment of publishing, so a revoked role stops the send.
Where a platform provides engagement APIs, numbers are real. Where it doesn't (personal LinkedIn), you can enter metrics manually and they're always labeled user-supplied. Performance signals nudge future drafts — with a visible explanation and an off switch.
14-day free trial. No credit card. Upload your archive, meet your style card.
Start free trial →