Lightspeed Core is currently working on updating their rag-content repository to Llama Stack v0.3.x. After this is completed we will need to update our rhdh-rag-content repo https://github.com/redhat-ai-dev/rhdh-rag-content to consume this. 

Once the RAG image is built you will need to use it alongside Llama Stack to see if it works on 0.3.x. You will update https://github.com/redhat-ai-dev/llama-stack/blob/main/Makefile#L16 and follow the README to set the permissions for the vector db. Then you will run Llama Stack like normal alongside LCS and run some queries, you need to see if there is rag chunks and/or referenced docs in the responses, you can also eye test the responses to see if they are pulling from the docs.

If you need to build and publish images to our Quay, navigate to https://github.com/redhat-ai-dev/rhdh-rag-content/actions/workflows/build-and-push.yml and hit 'Run Workflow'. Enter the docs # and make sure you hit experimental to avoid bugs making it into any stable image tags for 1.8.

For example, asking "Tell me how to setup the Config Map for Lightspeed Core in Lightspeed" should give you the whole yaml breakdown you'd expect.