Pipelines that just run.

DataFlow moves data between Postgres, ClickHouse, S3 and HTTP APIs on a schedule you control. One YAML file per pipeline, retries and alerts built in.

Read the docs

Declarative

Describe sources, transforms and sinks in a short config. DataFlow figures out ordering and batching.

Incremental

Cursor-based sync moves only new and changed rows, so nightly jobs finish in minutes, not hours.

Observable

Every run reports row counts, lag and errors. Failed batches retry with backoff and page you only when it matters.

Pipeline throughput dashboard — Throughput of a production pipeline over a week.