Assuming you have a working Kubeflow Pipelines installation, download docling_pipeline.yaml and the upload it to your Kubeflow Pipelines UI or via the CLI. Then, start a run of that pipeline to see the conversion process happen.

git clone https://github.com/bbrowning/docling-kfp-demo
cd docling-kfp-demo
source venv/bin/activate
pip install -r requirements.txt

python docling_convert_pipeline.py

This generates a docling_pipeline.yaml that you can now run in any Kubeflow Pipeline. This is a simple demo pipeline that downloads some PDFs, splits them into a handful of batches, and converts the batches across a cluster.