Kedro is an open-sourced Python framework for creating maintainable and modular data science code.

Kedro

We are trying to build an API using Kedro. I understand that the kedro loads data in a lazy manner. Is there any feature way we can persist this lookup data in DataCatalog? We might be able to keep a robust and fast API using a combination of lazy and eager load on DataCatalog.

Exposing Kedro with API seem to become more common, I have seen a few people asking about it recently, if you have more opinion about it, please share them with us and help us understand what is needed.
<https://github.com/kedro-org/kedro/issues/1846>

How are you exposing Kedro to your API? Are you using `session`  or just `DataCatalog`?

If you are using `session.run` - data are lazily loaded, and the dataset is released as soon as they are not needed.
If you are using `DataCatalog` directly - you have finer control to when to release or load the data