Sample Data

Note

Placeholder. Downloadable sample datasets are planned. The demo data described here already exists inside the platform.

The demo domain

The platform ships with a small, generated demo domain: a fictional shop with customers, orders and returns. It is deterministic (seeded), so every run and every test sees the same data, and it contains realistic personal fields so access policies can be demonstrated without real personal data.

Table	Layer	Arrives as	Notes
demo__customers	raw	API-style snapshot, merged on customer_id	Name and email are tagged confidential PII
demo__orders	raw	One day per partition, merged on order_id	Daily partitions from 2026-10-01
demo__returns	raw	CSV files dropped into landing/demo/	Each row records the file it came from
stg_demo__*	staging	Python (Polars) and dbt	Cleaned and typed
int_demo__order_returns	intermediate	dbt	Orders with their returns
customer_summary, daily_revenue	marts	pandas and dbt	Every column declared and described

Planned

• Downloadable snapshots of the demo tables (Delta and Parquet), with a data dictionary.

• A larger generated dataset for performance and burst-compute comparisons.

• A public open dataset loaded through the same ingestion path.