On this page
2 What this guide is about
Every maritime organization runs on documentation. The ISM Code mandates a structured Safety Management System. SOLAS requires dozens of certificates per vessel. IMCA M 103 demands detailed FMEA worksheets and DP Operations Manuals. MARPOL requires ongoing record-keeping across six pollution categories. A single company operating a fleet of twenty vessels might maintain thousands of interconnected documents, and a regulatory change at the IMO level must cascade through flag state legislation, classification society rules, company procedures, and vessel-specific manuals.
The current approach to managing this cascade is largely manual. Procedures are drafted in Microsoft Word, frozen into PDFs, and distributed via email or shared drives. When a regulation changes, someone must identify every affected document, open each one, make edits, re-approve, and redistribute. Cross-references between documents are static text; if the referenced document moves or changes, nothing alerts you. Version control means appending "v2_FINAL_reviewed_FINAL2" to a filename.
This guide introduces two ideas that, together, offer a way out. The first is the Abstract Syntax Tree, a way of representing any structured document as a tree of data that computers can read, validate, and transform. The second is Markdoc, a content authoring system built by Stripe that turns familiar Markdown writing into exactly this kind of structured data. Together, they power platforms like Manifest that treat maritime documentation not as static files but as living, queryable, composable data.
The narrative builds from foundations to applications. We start with what an AST actually is (it is simpler than it sounds), move through how Markdoc works and what makes it different, then explore why these concepts are particularly powerful for an industry defined by hierarchical, cross-referenced, regulation-driven documentation.