Hey @rnahasnahasuddin
Every AI Center project — whether created manually or via Document Understanding - is an isolated environment that holds its own ML packages, datasets, pipelines, and skills.
If you create a project directly in AI Center, it’s meant for managing custom ML workflows (like training or deploying models). But when a DU workflow triggers the creation of a project (e.g. by using an out-of-the-box ML Skill like Invoice Extractor), that project appears in AI Center too - yet it’s tied to the DU pipeline and uses prebuilt models.
They may both show up in the same AI Center UI, but their purpose and usage are different.