Credentials stay explicit — and local

Connect the data you already have

Databases, warehouses, object storage, files, and APIs — each integration is an optional install (pip install ciaren[postgres]) so the core stays lean.

Available today

Ships in the open-source core

SQL databases

PostgreSQL

[postgres]

Query results in, tables out — with custom SQL support.

MySQL / MariaDB

[mysql]

Read and write over the standard wire protocol.

SQLite

Built in

Zero-setup local databases. Works out of the box.

DuckDB

[duckdb]

In-process analytics database for fast local OLAP.

SQL Server

[mssql]

Microsoft SQL Server via ODBC.

Snowflake

[snowflake]

Read from and write to Snowflake warehouses.

Object & document storage

Amazon S3

[s3]

Buckets, custom regions, and S3-compatible endpoints (MinIO, R2…).

Azure Blob Storage

[azure]

Including custom endpoints for sovereign clouds.

Google Cloud Storage

[gcs]

Read and write GCS buckets.

MongoDB

[mongo]

Collections in and out of your flows.

Local folders

Built in

Your filesystem is a first-class source. No setup.

File formats

CSV / TSV

Built in

Streaming reads on the Polars engine.

Parquet

Built in

Columnar files, read and written natively.

Excel (.xlsx)

Built in

Spreadsheets in, spreadsheets out.

JSON / JSONL

Built in

Nested and line-delimited data.

Text

Built in

Plain-text files for simple pipelines.

APIs & automation

REST APIs

Built in

Read-only HTTP connector with auth, pagination, and SSRF protection.

Webhooks

Built in

Trigger any flow from external systems over HTTP.

REST API & CLI

Built in

Everything the UI does is a documented FastAPI endpoint; scripting via the ciaren CLI and Python client.

Cron scheduling

Built in

Built-in timezone-aware scheduler with retries and catch-up.

Machine learning

MLflow

Built in

Every training run tracked — local store, remote server, or Databricks.

scikit-learn

Built in

Classifiers, regressors, clustering, and PCA built in.

XGBoost & LightGBM

[ml]

Gradient-boosted models as optional extras.

Coming soon

Where integrations are headed

From the public roadmap — directions under active exploration, not dated commitments. Most new integrations will arrive as plugins.

Planned

Connector plugins

More databases, SaaS APIs, and storage backends delivered as marketplace plugins rather than core bloat.

Planned

Notebook export

Export flows as Jupyter notebooks alongside the existing pandas and Polars scripts.

Planned

Validation exports

Turn data-quality checks into external test and validation formats.

Planned

Job-template export

Reusable job templates for handing pipelines to orchestrators like Airflow.

Planned

AI providers

Assistive workflows through opt-in AI provider plugins — never a required dependency.

Planned

Richer automation triggers

Deeper REST, CLI, and webhook-based ways to run flows outside the UI.

Missing a connector you need? The plugin SDK lets you build it today — or open a discussion and make the case.

Explore the plugin SDK Request an integration

Your data, your machine, your pipelines

Install locally and connect your first database in minutes.

Get started on GitHub Read the Docs

AGPL-3.0No account requiredRuns on your machine