For indie authors who publish seriously

Draft in. Book out.

One AI-powered pipeline takes your manuscript through every production step — editing, formatting, metadata, ebook conversion, KDP setup — and delivers a publish-ready book. Includes a top-tier AI text detector built in.

See pricing See the pipeline
1
Configure
Set genre, formatting, pricing, and how much control you want to keep.
You
2
Edit
Structural review, line edit, copy edit, and proofread — all from one prompt.
AI agent
3
Format
Interior typography, front and back matter, ebook conversion, print PDF.
AI agent
4
Metadata
Blurb, keywords, categories, A+ Content, and a ready-to-paste KDP checklist.
AI agent
5
Publish
Upload the finished files, click publish, verify the listing. You're live.
You
The Pipeline

Ten phases. Zero gaps.

Every step from raw manuscript to live listing, in order. The agent handles what it can. You handle what it can't. Nothing falls through.

01
Manuscript QC
Word counts, structure, continuity, character audit, POV check, dialogue ratio.
Agent
02
Dev Edit
Plot arc, pacing map, character arcs, scene necessity, opening & ending strength.
Agent
03
Line & Copy Edit
Prose polish, voice, filter words, grammar, style sheet, fact-check.
Agent
04
Proofread
Final error pass. Formatting consistency. Chapter numbering. Manuscript locked.
Agent
05
Cover Design
Detailed brief with comps, mood, palette, specs. You bring the designer.
Human gate
06
Interior Format
Typography, trim, margins, front & back matter, chapter styling, print PDF.
Agent
07
Ebook Conversion
EPUB & KPF export. TOC, hyperlinks, device preview, file optimisation.
Agent
08
Metadata & Blurb
Keywords, categories, blurb variants, A+ Content, pricing, metadata package.
Agent
09
KDP Upload
Step-by-step checklist with exact values. You click the buttons.
Human gate
10
Launch
Listing verified. Series linked. Ads brief. Post-launch KPI schedule.
Agent
Clarity — built-in AI detector

24 patterns. 8 metrics.
Sentence-level precision.

Not a pass/fail checkbox. Clarity scans your manuscript for 24 known AI writing patterns across 3 severity tiers, plus 8 statistical metrics — and tells you exactly what to fix and how. Each scan checks 185 individual sub-patterns.

18/100
Reads human — low AI pattern density
No recognisable AI writing patterns. This reads like a person wrote it.
4/24 patterns detected · 1/8 statistical flags · Deep scan: 185 sub-patterns checked
Sentence length variance
1–55 words (avg 14)
← Natural writingAI-typical →
This measures how much sentence length varies. Human writing mixes short punchy sentences with long flowing ones. Here, sentences range 1–55 words — strong natural variation.
CV = 0.88 · 296 sentences measured
Dialogue variance
136 turns, avg 15 words (1–74)
← Natural writingAI-typical →
This detects "ping-pong" dialogue — the AI habit of alternating single short sentences between speakers. 63% of 136 turns are single sentences, with 6 short exchanges in a row. Slightly uniform.
Turn CV = 1.05 · 63% single-sentence · Longest short run: 6 · 38% multi-sentence
Lexical diversity
77% unique words
← Natural writingAI-typical →
This measures how many different words appear in any 100-word stretch. AI-generated text tends to recycle the same vocabulary. Here, 77% of words are unique — a rich, varied vocabulary.
TTR = 0.77 · Measured in sliding 100-word windows
Showing 3 of 8 statistical metrics · Full report includes all 8 + 24 pattern categories
◈
24 pattern categories, 185 sub-patterns
From "delve" and "tapestry" to throat-clearing intros, sycophantic openings, and conclusion mirroring. Three severity tiers per pattern. Nothing vague.
◉
8 statistical metrics
Sentence variance, paragraph uniformity, lexical diversity, opener repetition, passive voice, formal conjunctions, dialogue ping-pong detection, and more.
◧
Every hit explained and fixable
Each flagged pattern comes with a plain-English explanation of why it's a problem and a concrete suggestion for how to fix it. Not just a score — an editorial tool.
◇
Runs on full manuscripts
Most detectors choke on anything longer than a blog post. Clarity handles 80,000+ words and scales the statistical analysis correctly across book-length text.
◫
Annotated text output
The exact words and phrases that triggered each pattern are highlighted in your text, colour-coded by severity. You see precisely what to change.
⬡
Use it during drafting, not just after
Run Clarity on individual scenes while you're still writing. Feed the stats back into your editing process. Iterate until the scores drop. It's a revision loop, not just a final check.
Why Draft Kiln

Built from real production.

Every phase comes from producing real books across five genres. Nothing theoretical. Nothing guessed.

◈
Genre-aware defaults
Cozy mystery, Scandinavian noir, fantasy romance, sci-fi, literary — each gets purpose-built typography, layout, and editorial tone presets.
◉
You choose the autonomy
Full auto, checkpoint approval, editorial hold, or fully supervised. Decide how much control to keep at every phase.
◧
Not a black box
Every phase produces visible reports and deliverables. You see the word counts, the edit log, the pacing map. You approve before it moves on.
◫
Global-ready
US Letter and A4 trim sizes, USD / EUR / GBP pricing, imperial and metric units. Built for publishers on both sides of the Atlantic.
⬡
AI detector built in
Sentence-level AI text detection with five-signal analysis. Not a pass/fail checkbox — a full statistical breakdown you can export and act on.
◇
House rules enforced
Set scene card compliance, dialogue ratios, word count minimums, and editorial boundaries once. The agent follows them every time.
How it works

Four steps. One afternoon.

Set it up once per book. The system handles the rest.

1
Configure the job ticket
Use the web configurator to set everything — genre preset, formatting, pricing, autonomy level, back matter, metadata approach. Lock it when you're done. Export JSON.
2
Set up the production folder
Run the setup script. Drop your manuscript and the job ticket into the folder. That's the input.
3
Let the agent work
Point your AI tool at the folder, paste the processing prompt, and say go. The agent runs each phase, pauses at checkpoints and human gates, and saves deliverables as it goes.
4
Publish
When the pipeline finishes, your deliverables folder contains the EPUB, print PDF, DOCX, and a metadata package with exact values for every KDP field. Upload and go.
Job Ticket Preview
title "Murder at the Marché" author "Elodie Marais" genre cozy_mystery series "Canal & Croissant" #3 language en units imperial platform KDP format ebook + print ku_enrolled true currency USD price $4.99 trim 5 × 8 in body_font Crimson Text chapter_font Playfair Display font_size 11pt drop_caps true scene_break ornament autonomy checkpoints dialogue_ratio 45–55% min_scene_wc 1500 keywords agent_researched blurb_style hook → stakes → tease
Pricing

One book or one hundred.
Start where you are.

No subscription required. Run a single book for $19, or subscribe when you're publishing regularly.

One Book
$19

Full pipeline, one manuscript. No subscription.

  • All 10 pipeline phases what's this?QC, dev edit, line edit, proofread, formatting, ebook, metadata, and launch — the full production chain.
  • Opus-grade AI editorial what's this?Your manuscript is processed by the most capable AI model available — not a cheaper, less accurate substitute.
  • 1 AI detector scan what's this?Run your finished manuscript through the AI text detector once and get a full report.
  • eBook + print output
  • KDP-ready metadata
  • Email support
Get started
Standard
$29 /mo

For authors publishing regularly.

  • 3 books/month
  • All 10 pipeline phases
  • AI detector — 5 scans/mo
  • Basic genre presets what's this?Pre-configured settings for common genres — typography, tone, and formatting applied automatically.
  • Production dashboard what's this?A visual overview of all your books and their current pipeline status in one place.
  • Email support
Start trial
Professional
$59 /mo

Full toolkit for serious indie authors.

  • 6 books/month
  • All 10 pipeline phases
  • AI detector — unlimited
  • Full genre preset library
  • Job ticket configurator what's this?A web-based form to set every detail — trim, fonts, pricing, autonomy level — and export as a ready-to-use JSON file.
  • House rules engine what's this?Define editorial rules once — dialogue ratios, banned words, scene card compliance — and the AI enforces them on every book.
  • Priority support
Start trial
Publisher
$129 /mo

For indie houses and multi-author operations.

  • 15 books/month
  • 3 concurrent pipelines what's this?Process three books at the same time instead of waiting for one to finish before starting the next.
  • AI detector — unlimited
  • Multi-pseudonym management what's this?Manage separate author identities under one account. Each pseudonym gets its own settings, style, and editorial voice.
  • Custom house rule templates what's this?Create and save reusable rule sets per genre, per series, or per author. Apply them to new books with one click.
  • Series style libraries what's this?Lock in the formatting, fonts, chapter styling, and back matter for a series. Every new book matches automatically.
  • White-label export what's this?Remove all Draft Kiln branding from reports and deliverables. Present the output as your own.
Contact us
Ready?

Stop assembling.
Start publishing.

Run your first book through the full pipeline for $19. No subscription. No account to cancel. Just a manuscript in and a book out.

Run one book — $19 Compare plans