Private beta · waiting list open
Your AI editorial team, not an AI textbox.
Researchers, writers, editors and fact-checkers working through a real editorial pipeline — in your voice, backed by evidence, with you as editor-in-chief.
How it works
From idea to every channel.
Each stage has an owner, an output you can review, and a clear next step. Revise anything with your own instructions.
- 01
Idea
Capture angles and hypotheses worth writing about.
- 02
Research
Sources gathered, weighed by authority, contradictions flagged.
- 03
Brief
Audience, goal and key points agreed before a word is drafted.
- 04
Outline
Structure you can approve or redirect in one pass.
- 05
Draft
Written in your brand voice, following your editorial rules.
- 06
Review
Chief editor, SEO/GEO editor and fact-checker score the draft.
- 07
Publish
Apply the fixes you accept, then ship with confidence.
- 08
Repurpose
Turn one article into LinkedIn, X, newsletter and more.
The team
Nine specialists. One newsroom.
Every agent has a narrow job and strict rules, so the output is reviewable — not a wall of plausible text.
Researcher
Finds and ranks sources, extracts evidence, surfaces contradictions.
Writer
Turns research into briefs, outlines and drafts in your voice.
Chief Editor
Scores quality, structure and argument; lists prioritized fixes.
SEO / GEO Editor
Checks search and AI-citation readiness without keyword stuffing.
Fact-checker
Links every claim to evidence and flags what's unsupported.
Voice Coach
Learns how you write from samples and keeps drafts on-brand.
Strategist
Defines audience, pillars and differentiation for your niche.
Calendar Planner
Builds a dated plan with content mix and seasonal hooks.
Repurposer
Adapts finished pieces to each channel's native format.
Why it's different
Editorial standards, built in.
Evidence you can inspect
Every factual claim is mapped to sources with an authority tier, so you see what's strong, weak or contradicted before you publish.
Your voice, not generic AI
Train a voice profile from your own writing and set editorial rules — banned phrases, preferred terms, heading style.
You stay editor-in-chief
Agents propose; you approve. Pick which review findings to apply and revise any stage with your own instructions.
A calendar that plans itself
Generate monthly or quarterly plans with topic clusters, freshness updates and seasonal hooks, then drag to adjust.
One article, every channel
Repurpose into LinkedIn posts and articles, X threads, newsletters, Reddit discussions, video and podcast outlines.
Built for teams
Share projects with editors and viewers. Everyone sees the same research, versions and review history.
Formats
One pipeline, shaped to the format.
Each project picks a content type, and every agent adapts its structure, length and review criteria to it.
Blog articles
The full pipeline: research, brief, outline, draft, multi-agent review and repurposing.
LinkedIn posts
Short, hook-first posts drafted straight from research, reviewed for clarity and voice.
LinkedIn articles
Long-form thought leadership formatted for LinkedIn's reader, not a search engine.
Choose your models
Assign a provider and model to each agent — Anthropic, OpenAI, Google and more.
Your content stays yours
We never use your drafts or sources to train models. Keys are encrypted at rest.
Role-based access
Owners, editors and viewers — share projects without sharing everything.
FAQ
Questions, answered.
How do I get access?
We're onboarding teams gradually during the private beta. Join the waiting list by email and we'll reach out as soon as a spot opens.
Is this just another AI writing tool?
No. Instead of one text box, you get a team of specialised agents — researcher, writer, editors, fact-checker — working through an editorial pipeline you control. Every stage is visible and revisable.
Which AI models does it use?
Each agent can be assigned a model from providers such as Anthropic, OpenAI, Google and others. Your workspace administrator chooses the provider and model per agent.
Is my content used to train AI models?
We don't use your content to train models. Content is sent to the model providers configured for your workspace only to perform the tasks you request. See our Privacy Policy for details.
Can I keep my own writing style?
Yes. The Voice Coach learns from samples you provide, and editorial rules let you ban phrases, enforce preferred terms and set formatting conventions.
How does fact-checking work?
The fact-checker extracts claims from your draft and links them to evidence from research sources, rating each as strong, moderate, weak or unsupported so you know what to fix.
Can my team collaborate?
Yes. Project owners can invite editors (who can run agents and edit) and viewers (read-only). Sharing projects with teammates is part of the product roadmap.
Hire your editorial team soon.
We're onboarding teams gradually. Join the waiting list and we'll email you as soon as a spot opens.