The Secretary Suite Project
DOI: To be assigned
John Swygert
May 20, 2026
Abstract
A generation of living memory is passing rapidly into silence. Families, communities, professions, military units, churches, towns, creative circles, and private households contain vast unwritten archives of experience that are rarely preserved before death, illness, distance, or time makes recovery impossible. Traditional ghostwriting is expensive, slow, and inaccessible to many people. Generic artificial-intelligence writing tools, meanwhile, often produce flattened prose unless guided by a disciplined human-centered method.
This paper introduces the Human-AI Book Engine as a practical framework for transforming scattered human material into coherent, publishable manuscripts. The Book Engine is not a prompt trick, a replacement author, or a generic writing template. It is a structured collaborative system in which a human interviewer, author, family member, researcher, or project lead gathers testimony, asks precise questions, records the right details, builds a book outline, maintains a book bible, develops a secondary master document, storyboards the work with the AI agent, and repeatedly refines the manuscript through human judgment.
The central claim is simple: artificial intelligence becomes most valuable in book creation when it is used not to replace the human storyteller, but to preserve, organize, structure, and clarify human testimony at a scale and speed previously unavailable to ordinary people. In this model, anyone with discipline, care, and the willingness to ask the right questions can function as a practical ghostwriter for a family, a community, a legacy project, a memoir, a professional history, or a thematic book of lived experience.
Introduction
Most people die with books inside them.
They do not usually die with polished manuscripts. They die with fragments: stories told at kitchen tables, jokes repeated at reunions, memories hidden in photographs, military service rarely discussed, letters stored in boxes, recipes never written down, family grief carried privately, workplace knowledge absorbed over decades, and small acts of courage that nobody thought to record.
This is not merely a private loss. It is a cultural loss.
When a generation passes away, more disappears than names and dates. A living archive disappears. The way people spoke, worked, loved, fought, survived, forgave, raised children, built institutions, repaired things, worshiped, cooked, served, mourned, laughed, and endured disappears with them unless someone captures it.
The Human-AI Book Engine is designed for this problem.
It begins with a practical observation: artificial intelligence can write quickly, but speed alone does not make a book. A book requires structure, memory, theme, sequence, rhythm, emotional truth, factual grounding, reader orientation, and publishing logic. Without those elements, AI-generated writing often becomes generic, repetitive, emotionally flat, or structurally confused.
The Book Engine solves this by placing the human being back at the center. The AI does not become the author. The AI becomes the organizing instrument through which the author, family member, interviewer, researcher, or ghostwriter can transform scattered material into a coherent manuscript.
The engine depends on several essential components:
A clear premise.
A disciplined interview process.
A book outline.
A book bible.
A secondary master document.
Storyboarding between the human and the AI agent.
Voice calibration.
Publishing-package development.
Revision loops.
Series logic where appropriate.
If any of these pieces are removed, the system weakens. The Book Engine is not merely “ask AI to write a book.” It is a method for gathering human material, preserving its emotional and factual integrity, and converting it into readable form.
1. The Problem Of Vanishing Human Archives
Every family contains unwritten history.
A parent may know where the family came from. A grandparent may remember the war. A neighbor may remember a vanished town. A retired worker may understand how an industry actually functioned. A widow may carry the truth of a marriage that no official record can describe. A veteran may know what courage cost. A teacher may know what children used to need. A nurse may remember what suffering looked like before machines mediated everything. A craftsman may know techniques that were never written down.
Most of this disappears.
The disappearance is not always dramatic. It happens quietly. A person dies. A house is cleaned out. Letters are thrown away. Photographs lose their names. Children remember fragments but not chronology. Grandchildren inherit objects without stories. A family retains emotion but loses context.
Modern society is facing this problem at enormous scale. A large retirement-age generation is carrying memory from the twentieth century into the twenty-first. Much of that memory includes firsthand contact with older worlds: pre-digital childhoods, military service, industrial work, analog professions, religious communities, regional cultures, extended families, postwar economic change, and the transition into the internet age.
This material is historically valuable, but it is rarely treated as such while the people who hold it are still alive.
The traditional solution would be professional biography, memoir writing, oral history projects, or academic archiving. These remain valuable, but they are limited by cost, access, time, and expertise. Most families will never hire a professional biographer. Most elderly people will never be interviewed by an institution. Most communities will not receive formal archival attention.
The Human-AI Book Engine offers a practical alternative.
It allows ordinary people to become competent memory gatherers and manuscript builders. The family member does not need to be a professional ghostwriter. The friend does not need a literary degree. The community volunteer does not need a publishing staff. What they need is a structured system.
2. The Core Principle
The core principle of the Human-AI Book Engine is this:
The AI is not the author. The AI is the organizing instrument through which the author becomes more fully legible.
This distinction matters.
When artificial intelligence is treated as the author, the result often becomes hollow. The machine produces language, but the language may not carry the pressure of lived experience. It may sound polished while saying very little. It may create the appearance of a book without the moral, emotional, or factual weight of one.
When artificial intelligence is treated as an organizing instrument, the result can be very different. The human supplies memory, purpose, correction, judgment, preference, lived truth, anger, grief, humor, pacing, and approval. The AI helps sort, structure, draft, compare, summarize, expand, compress, and package.
This creates a working division:
The human provides the soul.
The AI provides structural acceleration.
The human remembers.
The AI organizes.
The human corrects.
The AI revises.
The human decides.
The AI assists.
That balance is the foundation of the engine.
3. The Book Engine Defined
The Human-AI Book Engine is a repeatable collaborative workflow for transforming raw human material into a publishable manuscript.
The raw material may include:
Recorded interviews.
Voice notes.
Text messages.
Family stories.
Letters.
Photographs.
Medical records.
Military records.
Professional experience.
Research notes.
Old blog posts.
Poems.
Songs.
Emails.
Journals.
Scattered memories.
Conversations.
Previously drafted fragments.
The engine receives this material and moves it through a structured sequence:
- Core premise.
- Emotional center.
- Reader promise.
- Book outline.
- Book bible.
- Secondary master document.
- Storyboard.
- Draft generation.
- Voice calibration.
- Revision loop.
- Publishing package.
- Series or archive expansion.
This sequence allows a project to move from chaos into form.
The engine is not rigid in the sense that every book must look the same. It is rigid in the sense that every serious project must answer certain questions before it becomes a book.
What is this book?
Who is it for?
Why does it matter?
What must not be lost?
What is the emotional center?
What is the structure?
What belongs in this book?
What belongs in a later book?
What facts must be checked?
What tone should the reader feel?
What is the final publishing form?
These questions prevent the manuscript from collapsing into fragments.
4. The Interview And Data-Gathering Layer
The Book Engine begins before the writing.
The first major task is gathering the right human material.
This requires more than asking, “Tell me your life story.” That question is too large. Many people cannot answer it well. They become overwhelmed, drift out of sequence, omit important details, or give generalized answers.
The interviewer must probe in layers.
Useful questions include:
Where were you born?
What did your childhood home look like?
Who was in the house?
What did your parents do?
What did the kitchen smell like?
What did Sunday feel like?
What was your first serious memory of fear?
Who taught you discipline?
Who taught you love?
Who disappointed you?
Who saved you?
What work shaped you?
What place formed you?
What loss changed you?
What mistake taught you?
What did you believe when you were young?
What do you believe now?
What do you want your family to understand after you are gone?
The goal is not merely information. The goal is recoverable life.
Dates matter. Places matter. Names matter. But emotion also matters. A correct date without emotional meaning is a record. Emotional truth without factual grounding is a mood. A book needs both.
The interviewer should gather:
Names.
Dates.
Places.
Relationships.
Events.
Turning points.
Repeated phrases.
Family sayings.
Objects.
Letters.
Photographs.
Lessons.
Contradictions.
Regrets.
Humor.
Faith.
Anger.
Silences.
The silences are important. When someone avoids a topic, hesitates, cries, laughs unexpectedly, or changes direction, the interviewer has reached a living edge. That edge may require gentleness, but it may also contain the heart of the book.
The Book Engine depends on this material. The AI can organize what is given. It cannot recover what was never asked.
5. The Core Premise
Before drafting begins, the project must identify its core premise.
The core premise is not the title. It is the central reason the book exists.
Examples:
This is a book about a father whose life demonstrates duty without arrogance.
This is a book about surviving grief after the event that breaks a life.
This is a book about a family business that carried a town through change.
This is a book about a woman whose private sacrifices held a family together.
This is a book about a soldier’s life after war.
This is a book about the emotional cost of caregiving.
This is a book about a community before it vanished.
The premise prevents drift. It tells the AI agent and the human operator what belongs inside the book and what belongs outside it.
Without a core premise, the manuscript becomes a storage unit.
With a core premise, the manuscript becomes architecture.
6. The Emotional Center
The emotional center is the living pressure behind the book.
Two books may share the same subject and still be completely different because their emotional centers differ.
A book about a father may be centered on gratitude.
Another may be centered on reconciliation.
Another may be centered on grief.
Another may be centered on justice.
Another may be centered on inheritance.
Another may be centered on warning.
The Book Engine must identify this early because it affects every later decision: title, chapter order, tone, pacing, cover design, subtitle, conclusion, and reader promise.
The emotional center should be stated plainly.
Examples:
The emotional center of this book is the desire to honor a quiet life of service.
The emotional center of this book is the struggle to live after tragedy.
The emotional center of this book is the preservation of a mother’s voice.
The emotional center of this book is the recovery of dignity after humiliation.
The emotional center of this book is the need to pass wisdom forward before it disappears.
The AI agent must be told this repeatedly throughout development. Otherwise, it may drift into generic book language.
7. The Reader Promise
A book must make an implicit promise to the reader.
The reader promise answers the question: Why should someone give this book their time?
In a legacy book, the promise may be intimate: this book will preserve a life so the family can remember.
In a public memoir, the promise may be broader: this book will help the reader survive loss.
In a professional history, the promise may be educational: this book will preserve knowledge from a field or institution.
In a philosophical book, the promise may be interpretive: this book will give the reader a new way to see human experience.
The Book Engine should define the reader promise before the manuscript is fully drafted.
A strong reader promise might be:
This book will help readers understand that an ordinary life, lived with discipline and love, can become extraordinary without ever seeking applause.
Or:
This book will help grieving readers understand that surviving the breaking is only the first part; learning to live through the after is the deeper work.
Or:
This book will show families how to preserve the stories of elders before those stories vanish.
The reader promise protects the book from becoming self-indulgent. It turns private memory into shared meaning.
8. The Book Outline
The book outline is the skeleton.
It defines the visible structure of the manuscript.
A weak outline merely lists topics.
A strong outline creates progression.
The outline should include:
Title.
Subtitle.
Part structure if needed.
Chapter numbers.
Chapter titles.
Chapter purpose.
Emotional movement.
Key material to include.
Approximate sequence.
The outline does not have to be perfect at first. It is a working map. But the AI agent must understand it before drafting substantial manuscript sections.
A useful outline entry might look like this:
05 Chapter
The Room Where The Story Changed
Purpose: This chapter introduces the central turning point in the subject’s life. It should begin with physical setting, move into the event itself, then show how the person changed afterward. The tone should be restrained, intimate, and observational. Avoid melodrama. Preserve concrete details.
This level of outline instruction gives the AI agent direction without imprisoning the writing.
The outline is not the book. It is the first form of the book.
9. The Book Bible
The book bible is the controlling document.
It contains the rules, facts, tone, structure, terminology, names, formatting conventions, and non-negotiable decisions for the project.
If the outline is the skeleton, the book bible is the memory system.
The book bible should include:
Correct title.
Correct subtitle.
Author name.
Series name if applicable.
Purpose of the book.
Target reader.
Tone rules.
Formatting rules.
Chapter heading style.
Recurring motifs.
People and relationship descriptions.
Correct spellings.
Chronology.
Important dates.
Sensitive topics.
What to include.
What to avoid.
Approved phrases.
Disallowed phrases.
Publishing notes.
Cover notes.
KDP positioning notes.
Category and keyword logic.
The book bible prevents inconsistency.
It tells the AI agent:
Do not rename the book.
Do not change the subtitle.
Do not alter chapter format.
Do not make the tone sentimental if the book requires restraint.
Do not overuse certain words.
Do not invent facts.
Do not include material that belongs in another book.
Do not violate the author’s house style.
This document is essential because AI systems can drift. They can rename sections, smooth over important emotional edges, repeat phrases, flatten tone, or reintroduce rejected language. The book bible keeps the project anchored.
Without a book bible, every session risks becoming a new book.
With a book bible, the project remembers itself.
10. The Secondary Master Document
The secondary master document is different from the book bible.
The book bible controls the rules.
The secondary master document collects the developing material.
It may contain:
Expanded notes.
Interview fragments.
Scene ideas.
Chapter fragments.
Unused passages.
Alternative titles.
Cut material.
Future book ideas.
Timeline pieces.
Emotional observations.
Research notes.
Quotes from the subject.
Lists of episodes to develop.
Questions still unanswered.
Publishing tasks.
Cover variants.
The secondary master document is the working reservoir.
It protects valuable material from being lost just because it does not fit the current chapter. It also prevents the main manuscript from becoming overcrowded.
This is especially important in legacy, memoir, grief, family, and research-based books because the material often arrives out of order. A person remembers something from childhood while discussing old age. A photograph triggers a story that belongs three chapters earlier. A phrase appears that may later become a chapter title. A side story may not belong in the current book but may become the seed of a sequel.
The secondary master document catches all of it.
Without this document, the human operator and AI agent must either force everything into the manuscript or risk losing valuable material. Both outcomes are bad.
The secondary master document creates breathing room.
11. Storyboarding With The AI Agent
Storyboarding is the stage where the human and AI agent decide how the book should move.
This is not only for visual media. Books need storyboards too.
A storyboard may define:
Opening image.
Chapter sequence.
Emotional arc.
Key scenes.
Transitions.
Recurring symbols.
Where to place letters or photographs.
Where to slow down.
Where to compress.
Where to let the reader breathe.
Where to insert humor.
Where to use silence.
Where to end a chapter.
Where to split a book into two volumes.
Storyboarding is especially important when the source material is emotionally intense. Grief, trauma, family history, military service, illness, caregiving, and personal survival can overwhelm a manuscript if not sequenced carefully.
The human operator may know the facts but not yet know the best order.
The AI agent may suggest possible structures:
Chronological.
Thematic.
Two-part before-and-after.
Braided timeline.
Memory-based.
Lesson-based.
Object-based.
Letter-based.
Interview-based.
Place-based.
The human then judges what feels true.
The storyboard is where the book becomes readable.
12. Voice Calibration
Voice calibration is one of the most important parts of the Book Engine.
AI writing often fails because it sounds like AI writing.
It may be too smooth.
Too inflated.
Too sentimental.
Too repetitive.
Too abstract.
Too polished in the wrong way.
Too eager to explain.
Too generic.
Voice calibration corrects this.
The human operator must tell the AI agent how the book should sound.
For example:
Restrained, not dramatic.
Warm, not sugary.
Direct, not academic.
Elevated, not pompous.
Funny in small flashes, not comedic throughout.
Precise, not sterile.
Emotional, not sentimental.
Angry energy translated into formal prose.
Plainspoken, but not simplistic.
The voice must match the project.
A father-honoring book may require humility, restraint, dignity, and gratitude.
A grief book may require tenderness, honesty, and emotional clarity.
A technical paper may require precision and conservative claims.
A comic book may require timing, absurdity, and sharpness.
The AI agent must be corrected whenever it drifts. Over time, the human and AI develop a project-specific working voice.
This is not magic. It is repeated correction.
The human says:
Too soft.
Too generic.
Too much.
Not enough emotion.
Do not use that phrase.
Keep the anger but make it publishable.
Make it cleaner.
Make it less flowery.
Make it more direct.
Preserve the wound.
Do not flatten this.
That loop is the engine learning the project’s working standard within the active collaboration.
13. Draft Generation
Once the premise, outline, book bible, secondary master document, storyboard, and voice calibration are in place, drafting can begin.
The AI agent should not be asked to write the entire book blindly.
Instead, drafting should proceed in controlled sections:
Title page.
Contents.
Prologue.
Chapter 1.
Chapter 2.
Part transitions.
Conclusion.
Back cover material.
KDP description.
Each drafted section should be checked against:
The outline.
The book bible.
The emotional center.
The reader promise.
The factual record.
The intended tone.
The project’s formatting rules.
The human operator must remain active. The AI can produce a strong draft, but the human must decide whether it is true.
A draft is not finished because it is grammatically correct.
A draft is finished when it carries the right meaning in the right structure with the right voice.
14. The Revision Loop
The revision loop is where the Book Engine becomes powerful.
The first draft is rarely the final form. The human operator reviews, reacts, corrects, and sharpens.
Typical revision commands include:
Make this less repetitive.
Make this more emotionally honest.
Cut the generic language.
Strengthen the opening.
Preserve the detail but reduce the explanation.
Move this idea earlier.
This belongs in the secondary master document.
This chapter needs a better ending.
This sounds too much like a speech.
This needs to sound more like the subject.
This is true, but it is in the wrong place.
This paragraph is the heart of the chapter; build around it.
This process is not a failure of AI. It is the proper use of AI.
The revision loop is where human judgment guides machine output into manuscript form.
A professional book does not emerge from one command. It emerges from structured iteration.
15. Publishing Package
The Book Engine does not end at the manuscript.
A book must also be positioned.
This requires a publishing package.
The publishing package may include:
Final title.
Subtitle.
Series title.
Volume number.
KDP description.
Rear cover small blurb.
Rear cover large blurb.
Author bio if needed.
Keywords.
Categories.
Cover text hierarchy.
Trim-size logic.
Hardcover, paperback, and ebook pricing strategy.
Series consistency.
Search behavior.
Reader expectation.
The publishing package matters because readers do not encounter the manuscript first. They encounter the cover, title, subtitle, description, category, and price.
A strong book can fail if it is packaged badly.
The Book Engine therefore treats publishing materials as part of the work, not as an afterthought.
For example, a two-book grief series must make the relationship between Book One and Book Two immediately clear. The title fields, subtitles, cover hierarchy, and keyword strategy must help readers understand that the books belong together while each volume has its own focus.
This is not cosmetic. It is reader orientation.
16. Series Logic
Many books do not stand alone.
One book may create another.
A memoir may lead to a workbook.
A family history may lead to a letters volume.
A technical paper may lead to a booklet series.
A grief book may split naturally into the breaking and the after.
A father-honoring book may lead to a broader book about duty, service, and family.
The Book Engine should identify this early.
Series logic asks:
Is this one book or several?
What belongs in Book One?
What belongs in Book Two?
Should the subtitle carry the distinction?
Should the covers match?
Should the keywords differ?
Should the volumes be priced the same?
Should the reader be able to understand the sequence immediately?
Series logic prevents overcrowding. It allows each book to breathe.
It also protects the author from trying to force an entire life, theory, family, archive, or philosophy into one volume.
17. The Human Operator As Practical Ghostwriter
One of the most important implications of the Book Engine is that ordinary people can become practical ghostwriters.
This does not mean everyone becomes a great literary artist automatically.
It means that a disciplined person can now do work that once required far more money, time, and specialized infrastructure.
A son can interview a father.
A daughter can preserve a mother.
A grandchild can record a grandparent.
A church can preserve the memory of elders.
A town can collect local stories.
A veteran group can record service histories.
A family business can document its origin.
A caregiver can preserve the story of illness and devotion.
A retiree can convert professional knowledge into a book.
The key is disciplined inquiry.
The human operator must learn to ask:
What happened?
When?
Where?
Who was there?
What did it feel like?
What changed afterward?
What did you learn?
What do you wish people understood?
What should not be forgotten?
The AI agent then helps organize the answers into a book-shaped form.
This democratizes ghostwriting.
It does not eliminate the need for care. It increases the need for care. The easier it becomes to generate text, the more important it becomes to gather truth.
18. Ethical Guardrails
The Human-AI Book Engine requires ethical discipline.
The AI agent must not invent facts.
The human operator must not falsely improve a life by removing complexity.
Sensitive material must be handled carefully.
Living people must be treated responsibly.
Private trauma should not be exploited.
Memory should not be presented as verified fact when it remains uncertain.
Family disputes should be contextualized.
Medical, military, legal, and professional claims should be checked when possible.
The subject’s dignity must be protected.
When uncertainty exists, the manuscript should say so.
Examples:
The family remembers this event as occurring in the early 1960s, though the exact year remains uncertain.
The author recalls the conversation clearly, but no written record of the exchange has been located.
Several relatives remembered the event differently; this version reflects the account most consistently repeated.
This kind of honesty strengthens the work.
The Book Engine should preserve memory, not manufacture legend.
19. Why This Matters Now
The need for this system is urgent.
Families are geographically scattered. Attention spans are fractured. Digital storage is chaotic. Elderly relatives may have photographs on phones, letters in boxes, records in drawers, and memories nobody has asked them to explain.
Meanwhile, AI tools are becoming widely available, but many people do not yet understand how to use them well.
The danger is that people will either ignore the opportunity or misuse it.
They may produce generic AI memoirs that sound polished but empty.
Or they may fail to preserve anything at all.
The Human-AI Book Engine offers a third path.
It says:
Do not ask AI to replace the person.
Ask the person better questions.
Record the answers.
Build the outline.
Create the book bible.
Maintain the secondary master document.
Storyboard the manuscript.
Draft carefully.
Revise repeatedly.
Package professionally.
Preserve the voice.
Publish the memory.
This process can help families, independent authors, community historians, caregivers, veterans, churches, small businesses, artists, and researchers transform human material into durable form.
20. The Book Engine As Memory Preservation At Scale
The larger significance of the Book Engine is memory preservation at scale.
A single professional biographer can only write so many books.
A single academic archive can only preserve so many collections.
But millions of ordinary people, equipped with a disciplined AI-assisted method, could begin preserving the memories closest to them.
This could create:
Family legacy books.
Community histories.
Retirement memoirs.
Caregiving memoirs.
Veteran testimony collections.
Marriage and family archives.
Professional wisdom books.
Local history volumes.
Grief and healing books.
Creative project books.
Life lesson books.
The Book Engine makes the process more accessible.
It lowers the barrier between memory and manuscript.
It allows a family to move from “someone should write this down” to “we are writing this down.”
That shift matters.
21. Limitations
The Book Engine is not a substitute for truth.
It cannot recover details nobody remembers.
It cannot verify claims without records.
It cannot replace the moral judgment of the human operator.
It cannot make an unwilling subject honest.
It cannot turn careless questioning into deep testimony.
It cannot solve every legal, ethical, or publishing concern.
It also cannot guarantee literary greatness.
What it can do is provide structure, speed, continuity, and organization. It can help a human being gather, shape, and preserve material that might otherwise remain scattered or vanish entirely.
The Book Engine is therefore best understood as a practical framework, not a miracle machine.
Conclusion
The Human-AI Book Engine is built on a simple but powerful idea: artificial intelligence should not erase the human author; it should help the human author become readable, organized, preserved, and publishable.
At its best, the engine allows scattered human material to become a coherent book. It gives families a way to preserve elders. It gives survivors a way to structure grief. It gives independent authors a way to develop manuscripts efficiently. It gives communities a way to record memory before it disappears.
The essential components cannot be skipped.
The project needs a core premise.
It needs an emotional center.
It needs a reader promise.
It needs a book outline.
It needs a book bible.
It needs a secondary master document.
It needs storyboarding.
It needs voice calibration.
It needs revision.
It needs a publishing package.
Together, these components form a practical engine for transforming lived experience into durable written form.
The final purpose is not merely efficiency.
The purpose is preservation.
A person’s life should not disappear because nobody knew how to organize the stories. A family’s history should not vanish because nobody could afford a ghostwriter. A generation’s memory should not be lost because the tools arrived too late or were used too shallowly.
The Book Engine offers a method for doing the work now.
Ask the questions.
Record the answers.
Build the structure.
Preserve the voice.
Make the book.
Note
A companion paper, “The People’s LLM Book
Engine,” extends this framework into a public
guide for ordinary users who want to use
ChatGPT and other large language models to
write, organize, publish, and preserve their own
Work.
References
Swygert, John. The Human-AI Book Engine Working Method. Unpublished collaborative development notes, 2026.
Swygert, John. After The Breaking. Ivory Tower Publishing, 2026.
Swygert, John. After The Breaking: Book II — Living Through The After. Ivory Tower Publishing, 2026.
Swygert, John. The Bridge To The Long Gray Line: A Father, A Family, And A Life Not Wasted. Ivory Tower Publishing, 2026.