// the find
elyra-ai/elyra
Elyra extends JupyterLab with an AI centric approach.
Elyra is a JupyterLab extension suite that adds a visual pipeline editor, batch job submission, and code snippet management to your notebook environment. The pipeline editor targets data scientists who want to wire up Kubeflow Pipelines or Airflow DAGs without writing orchestration code. It fits best in enterprise ML platforms (OpenDataHub, Kubeflow) rather than solo laptop setups.
The visual pipeline editor compiles to real Kubeflow Pipelines or Airflow DAGs — it's not a toy, it generates deployable artifacts. The metadata service for storing runtime configs, images, and component catalogs is genuinely useful for teams sharing a JupyterHub. Cypress E2E tests and CodeQL analysis show more engineering discipline than most JupyterLab extensions. Active maintenance as of May 2026 with Python 3.10+ and JupyterLab 4.x support.
The dependency surface is enormous — a full install pulls in Airflow, KFP SDK, LSP, nbdime, and more; version conflicts are a recurring issue in the GitHub tracker. The visual pipeline editor has a steep learning curve for runtime configuration: you need object storage (MinIO/S3), a runtime environment, and correct container images all wired up before a single node runs. Only ~2k stars for a project this mature suggests the JupyterLab extension ecosystem hasn't broadly adopted it. The Kubeflow Pipelines support lags behind KFP SDK v2's IR-based compilation model, which means you may hit compatibility walls on newer clusters.