Note that the logic to retrieve within the vector database and inject details in to the LLM context could be packaged while in the model artifact logged to MLflow working with MLflow LangChain or PyFunc model https://lilianrenz025886.wikiannouncement.com/user