All Tools
K
DataFreeOpen Source
KEDRO
Production-ready data science with modular, versioned ML pipelines
Apache-2.0
ABOUT
Data science projects often start as exploratory notebooks that are difficult to productionize — tangled code, no versioning, and manual data handoffs between steps. Kedro solves this by providing a framework that enforces modular pipeline architecture, automatic data cataloging, parameter configuration, and pipeline versioning — letting data scientists write production-ready code from day one without becoming software engineers.
INSTALL
pip install kedroINTEGRATION GUIDE
1. Build a modular, versioned data pipeline for training AI models with automatic data cataloging
2. Create reproducible ML experiments with parameterized pipelines that track data lineage across runs
3. Deploy a production data science workflow that runs in CI/CD with automated testing and validation
4. Implement a data transformation pipeline that loads from multiple sources, processes, and serves to ML models
5. Structure a team data science project with consistent conventions for code, data, and configuration management
TAGS
pipelinedata-sciencemlopsworkflowversioningpythonreproducibilityproduction