About the Companion Book
This documentation and the book Research Data that Can Be Trusted describe the same open-source platform and are written by the same authors. The two are designed to complement each other: the documentation is self-contained and free — every page can be read, and every example can be run, using only the code, configuration and synthetic data in this repository — while the book offers an extended conceptual treatment of the ideas behind the platform, including data provenance, the regulatory context for health data, and a taxonomy of data transformations. The book is optional enrichment; it is never a prerequisite for anything in this documentation.
Bibliographic information
Title: Research Data that Can Be Trusted
Authors: Bouzinier et al.
Series: SpringerBriefs in Computer Science
Publisher: Springer, 2026
Link: https://tidd.ly/4y1ClDH
Scan the QR code (or follow the link above) to get the book.
Chapter-to-documentation map
Each book chapter that describes the platform (Chapters 5–8 and the appendices) has a single canonical landing page in this documentation; the earlier chapters cover conceptual background (provenance, the regulatory context, a taxonomy of transformations) with no direct code counterpart:
Book chapter |
Documentation page |
|---|---|
Ch. 5, “Language Design” |
|
Ch. 6, “Proof of Concept Implementation” |
|
Ch. 7, “Sample Application: Building ML-Ready Datasets” |
|
Ch. 8, “Dorieh Medicare Claims Data Pipeline” |
|
Appendix A (core YAML DSL syntax) |
|
Appendix B (DSL extensions) |
Using this documentation without the book
If you do not have the book, the following reading order covers the same ground end to end:
Read Why a data platform for the motivation behind the platform, then Concepts: the Dorieh approach for the vocabulary and design ideas used throughout the documentation. (If you already know the motivation, start directly with Concepts.)
Work through the climate tutorial, a runnable Bronze–Silver–Gold pipeline built on open data.
Follow the Medicare tutorial, a guided path through the same patterns applied to a production-scale health data pipeline (runnable with synthetic data), with the Medicare case study as its reference.
Consult the DSL reference — Data modeling and Data modeling extensions — when you write your own data models.