Fig. 1 · Interactive · Data pipelines / ingestion

dlt pipeline sandbox

Poke real dlt concepts (write dispositions, schema contracts, incremental loading) and watch the consequences land in DuckDB. The tables below keep real state between runs, so run it a few times.

Source
shop.orders
Destination
duckdb
Presets
B · Pipelineextract → normalize → load
REST API
extract
normalize
load
— run the pipeline to see what happens —
shop.orders0 rows
empty — no load has run yet
shop.orders__items0 rows
empty — no load has run yet
C · Takeaway