Exam Certified Generative AI Engineer Associate topic 1 question 41 discussion

We need to first vectorize the user question, these vectors will then be used to search the database for relevant documents which will create an augmented prompt that the response-generating LLM will use to provide an answer

A. 1. Embedding model → 2. Vector search → 3. Context augmentation prompt → 4. Response generation LLM

In a typical RAG-enabled chatbot, the process usually follows these steps: Embedding Model: Converts the user’s question into a vector representation. Vector Search: Finds relevant information based on the vector representation. Context-Augmented Prompt: Combines the retrieved information with the original question to create a prompt. Response-Generating LLM: Generates the final response based on the context-augmented prompt. So, the correct sequence is: A. 1.embedding model, 2.vector search, 3.context-augmented prompt, 4.response-generating LLM Option C suggests starting with the response-generating LLM, which doesn’t align with the typical RAG workflow. The LLM needs the context-augmented prompt to generate a relevant response, which is why it comes last in the sequence.