PipelineWise

PipelineWise is a Python framework for configuring, running, and operating Singer-based ELT pipelines. It moves source data into analytical destinations with log-based, incremental, or full-table replication and optional native bulk transfer through FastSync.

Important

Project scope

Available sources are MariaDB and PostgreSQL; available targets are PostgreSQL and Snowflake. Other packaged connectors, including the Snowflake source, are experimental. pipelinewise init also generates some legacy templates that are not packaged. Review Connectors before selecting a source and target.

Release v0.64.1 is the last version from before the connector set was reduced. It is a historical reference, not a recommendation to deploy an older release.

Start here

  1. Installation — install PipelineWise, preferably with Docker.

  2. Create a pipeline — configure an available source-to-target route.

  3. Run a pipeline — validate, import, run, and inspect a pipeline.

  4. Troubleshooting — diagnose failures and recover safely.

Core capabilities

Capability

Behaviour

Singer replication

Log-based change capture, key-based incremental loads, and full-table snapshots.

FastSync

Native FullSync and PartialSync transfers for supported database routes.

Schema evolution

Detects source schema changes and applies compatible target changes.

Load-time protection

Masks, hashes, or removes sensitive values before target loading.

Data-diff

Performs bounded aggregate reconciliation with auditable coverage and remediation.

Configuration as code

Generates connector JSON, catalogs, and state from version-controlled YAML.

Documentation