About

This section covers DATAMIMIC's design philosophy, its origins, and where to go next in the documentation.

Design Philosophy

DATAMIMIC is built on a model-driven, deterministic-first approach. Models are written in an XML DSL and can be composed from many fragments via <include>, fragment <param> declarations, and per-call <property> overlays. The same seeded model produces the same output across runs and environments. The processing core is built in Python and Rust on top of open standards.

Two editions, one DSL

• Community Edition (CE) — Python library and datamimic CLI. Developers and scripts call the library directly.
• Enterprise Edition (EE) — on-prem platform built around the same DSL: a Web UI, a FastAPI server, a Postgres-backed scheduler, Celery workers that encapsulate the EE Core engine, and an optional LSP service for in-editor authoring support. Generation is triggered from the UI or from CI via REST and dispatched to workers as tasks.

A model written for CE runs unchanged inside EE workers; EE adds the surrounding platform (governance, RBAC, audit, scheduling, UI, deployment) around it.

Origins

DATAMIMIC grew out of rapiddweller's test data consulting work with large and mid-sized enterprises in regulated industries. The platform encodes practices developed over years of engagements — schema modelling, referential integrity across systems, anonymisation, and audit requirements — into a product that engineering teams can run on their own infrastructure.

More to Explore

Continue with the Tutorial for a worked end-to-end example, or jump straight to the Reference for model, generator, and API details.