Deployment for python project using generative ai foundational models in ai launchpad

I have created a ipynb file implementing RAG with HANA. For local development I am using service keys from Ai core after setting the same as environment variables on my machine. 

Now, how do i deploy the same to BTP?

Do i deploy as cf application? if yes, is there any reference on how to provide ai core as a service dependency?

Do i deploy in ai core using workflows? if yes, is there any reference on how to provide ai core as a service dependency?

Thanks

Kayur

  @carlosbasto