Use it when
- You want to create reproducible ML pipelines for production.
- You want a fault-tolerant platform with built-in fault recovery functions.
- You want an open-source Kubernetes-native workflow automation platform.
- You want a cloud-agnostic framework expandable with other tools.
- You want Python, Java, and Scala SDK support.
- You want a platform that natively understands data flow between tasks.
- You want a reliable platform in edge case deployments and large-scale workflows.
- You want support for DAG or Steps based declaration of workflows.
Example stacks
Airflow + MLflow stack
Installation
pip install flytekit