Booklet Engine V1 process illustration: validated source, semantic model, composition, PDF generation and finalization, preflight, visual review, acceptance.
Booklet Engine is the production workflow behind my independent publishing practice. The aim is to make publication work reproducible, inspectable, and recoverable while preserving the human editorial decision.
The problem
A finished-looking PDF is not enough. Source wording, document structure, layout decisions, build dependencies, and verification evidence must remain understandable after the conversation or tool session ends.
The workflow
Validated source is converted into a semantic model, then passed into layout and composition. ReportLab generates the PDF and pypdf handles finalization. Preflight checks and page rasters support visual inspection before acceptance and promotion.
Design decisions
Content and layout have separate responsibilities. The engine can run offline; AI assistance is optional upstream. Technical checks inform editorial review, and the owner decides whether the publication meets its intent. Keeping source, decisions, and evidence together makes later revisions and recovery practical.
The result
The accepted V1 workflow is producing finished books for ViridianMint Publishing (previously Phoenix Rising Publishing). It supports a repeatable sequence from validated inputs to an accepted publication. The accompanying graphic is a process illustration, not a screenshot of an application interface or a claim of measured performance.