Editorial onboarding

Your book starts with the right context.

Structured editorial onboarding designed to capture the author's ideas, voice, constraints and references before writing begins.

Start my book Login

21: Guided brief sections
4: AI roles in the pipeline
4: Export formats
3: Interface languages

See the pipeline at work

Writer, Editor, Fact Checker and Judge: how each model's output becomes the next one's input, and how criticism feeds back until the book earns its score.

From idea to manuscript, step by step

The studio walks you through every stage — nothing is generated until your context is complete and confirmed by you.

1
Create your project

Start from scratch or import a complete project from a JSON file. Every book gets its own ticket and workspace.
2
Build the editorial brief

21 guided sections: thesis, audience, style rules, rigor, exclusions, visual elements and more — plus a chapter planner with a brief per chapter and a book bible for consistency.
3
Add voice and sources

Upload your writing samples, stylistic references and source material. Text is extracted automatically and your voice is never mixed with your references.
4
Configure and approve

Pick the model for each role, the quality threshold, fact checking and visual elements. Review the transparent cost estimate before anything runs — with your own API keys and a hard spending cap.
5
Multi-LLM generation

Writer, Editor, Fact Checker and Judge work chapter by chapter with revision loops, editorial memory and a final whole-book coherence check.
6
Visuals resolved in place

AI images with your chosen engine, data charts drawn from the chapter's own numbers and native tables — all placed where the text calls for them.
7
Review, edit and export

Polish everything in the interactive book editor, accept or request revisions per chapter, and export to PDF, DOCX, EPUB and Markdown.

A pipeline that criticizes itself

One model writing alone gets tired, drifts and flatters itself. Here four specialized roles check each other: each output becomes the next role's input, and criticism feeds back until the chapter earns its score.

✍️
Writer

Writes each chapter in YOUR voice, from your brief and the running editorial memory.
🔍
Editor

A different vendor corrects grammar, coherence and brief compliance — without rewriting your voice.
🔎
Fact Checker

Validates dates, names, numbers and cross-chapter consistency; flags what the material cannot support.
⚖️
Judge

A third vendor scores 1-10 against your brief and issues discrete, actionable corrections.
🔁
Revision loop

Below your threshold, the corrections go back to the Editor — up to your chosen number of rounds.

…and when every chapter has passed, a final whole-book evaluation hunts for contradictions, broken continuity and unkept promises across the entire manuscript.

🎛️ Variant: your models per role

Assign Claude, OpenAI or Grok to each role. Different vendors judging each other means no model grades its own homework.

🌐 Variant: web-verified facts

For technical or historical books, the Fact Checker can verify external claims against live web sources before the Judge scores the chapter.

🎨 Variant: visual elements

With visuals enabled, the Writer plans images, charts and tables in place, the Judge scores their fit, and your chosen engine renders them after approval.

✅ Quality is measured, not assumed

Every chapter carries an explicit score against YOUR acceptance criteria — you set the bar, the pipeline reaches it or tells you honestly.

🧠 Consistency across chapters

An editorial memory carries names, facts and timeline forward, and the whole-book evaluation catches repetition and contradictions a per-chapter view would miss.

🚧 Failures are contained

A stubborn chapter is parked and flagged for your review while the rest of the book keeps advancing — one weak unit never sinks the manuscript.

Everything the studio does

A complete production workspace — from the first idea to the exported book.

📋 Guided editorial brief

21 structured sections with auto-save, progress tracking and import/export of the whole project as JSON.

🗂️ Chapter planner & book bible

Plan every chapter with its own objective, key points and acceptance criteria; keep names, facts and terminology consistent with the book bible.

🎙️ Your voice, protected

Author samples and reference authors live in strictly separate blocks: the pipeline imitates YOUR voice and never copies your references.

🤖 Configurable multi-LLM pipeline

Choose the provider and model for each role — Claude, OpenAI or Grok — with a recommended one-click preset and your own quality threshold.

🔎 Fact checking, optionally web-backed

An independent checker validates dates, names, numbers and cross-chapter consistency — and can verify external claims against live web sources.

⚖️ An independent judge

Every chapter is scored 1-10 against your brief. Below your threshold, concrete revision instructions feed back until it earns its score — then the whole book is checked for coherence.

🖼️ Images, charts and tables

Opt in per book and pick the image engine (Gemini or OpenAI). Deep, brand-consistent image prompts; charts drawn deterministically from real chapter data; native tables in every export.

✏️ Interactive book editor

Edit chapter text with full version history, regenerate any image by editing its prompt, keep every version and choose your favorite — nothing is ever deleted.

💰 Transparent cost control

Bring your own API keys (stored encrypted), see the over-estimated cost before starting, and rely on a hard per-book spending cap that stops generation before overspending.

📦 Professional exports

Download the whole manuscript or single chapters as PDF, DOCX, EPUB or Markdown — with your selected images, charts and tables embedded.

🔁 Review workflow

Accept chapters or request revisions with comments; delivery is never silent about content that stayed below your quality bar.

🌐 Three languages, your content language

Interface in English, Spanish and Italian; your book is written in the content language you choose and never auto-translated.

Why authors choose Ingenarte

🧭 You direct, the pipeline executes

Nothing is generated until your context is complete and you approve the cost. Every setting — models, threshold, revisions, visuals — is yours to decide.

🗣️ A book that sounds like you

The writer imitates your authentic samples, not a generic AI voice — and the judge penalizes any drift from your brief.

🛡️ No silent failures

A chapter that cannot reach your quality bar is flagged for review, never hidden — and it never blocks the rest of the book from being delivered.

📚 Nothing is ever lost

Immutable chapter versions, full image history, audit trail of every step and daily backups. You can always go back.

What the author thinks

Thesis, audience and argument captured as a structured brief.

How the author writes

Authentic samples and stylistic references, always kept separate.

How the book should be produced

Explicit rules, rigor and acceptance criteria for every chapter.

Ready to give your book the context it deserves?

Create your project, complete the guided brief and let the editorial pipeline do the heavy lifting.

Start my book