AI0-001 Implementing AI Solutions Practice Question

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

• ✗

  Vector store index

  Why it's wrong here

  A vector store index holds and queries embeddings; it does not create them. Embeddings come from a separate embedding model applied to each chunk. Indexing is the right choice when you already have vectors and need nearest-neighbour retrieval at scale, not when text must first be converted.

• ✗

  Document chunker

  Why it's wrong here

  The chunker splits source documents into smaller passages; it performs no numerical encoding. Conversion into vectors is the embedding model's job. Chunking would be correct when preparing long documents for retrieval, since chunk size and overlap affect what gets embedded, but it precedes embedding rather than performing it.

• ✗

  Large language model (LLM)

  Why it's wrong here

  An LLM generates text from a prompt; it does not produce the fixed embedding vectors used for similarity search. Embedding models map chunks into a shared vector space. An LLM is the right component when the retrieved context must be synthesised into a natural-language answer.

• ✓

  Embedding model

  Why this is correct

  Embedding models transform each document chunk into a dense vector, capturing semantic meaning so that similarity search can compare vectors via distance metrics. This satisfies the stem's requirement for numerical representations enabling retrieval, distinct from the LLM that generates answers or the vector store that indexes them.