mediumMultiple Select

Generative AI Leader Practice Question: Using Vertex AI RAG Engine to ground a chatbot in…

A company is using Vertex AI RAG Engine to ground a chatbot in internal documents. The chatbot sometimes returns outdated information. Which TWO steps should they take to improve freshness?

⚠ Common exam trap

The trap is confusing 'freshness' with other RAG tuning knobs (chunk size, temperature) — candidates who don't distinguish retrieval-index currency from generation parameters pick B or C.

Answer choices

Why each option matters

Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.

Correct answer & explanation

✓

Set up automated re-indexing on a schedule (e.g., daily)

Option A is correct because RAG Engine only retrieves from the indexed corpus, so scheduling automated re-indexing (e.g., daily via a pipeline or scheduled job) ensures newly updated or replaced documents are reflected in the vector index and stale chunks are removed. Option D is correct because attaching metadata such as version or timestamp to chunks lets retrieval filter or rank by recency, so the chatbot prefers the latest document version instead of matching an outdated chunk with similar embeddings. Option B is not appropriate because shrinking chunks to 50 tokens fragments context and does not address index staleness. Option C is wrong because raising temperature to 1.0 increases randomness and does not improve factual freshness. Option E is wrong because disabling grounding removes the internal-document retrieval entirely and falls back to potentially outdated pre-training knowledge.

Answer analysis

Option-by-option breakdown

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

✓
Set up automated re-indexing on a schedule (e.g., daily)
Why this is correct
Scheduled automated re-indexing periodically ingests updated source documents into the RAG corpus, so newly revised content replaces outdated embeddings. This directly satisfies the freshness requirement by preventing the index from serving stale chunks after source documents change.
✗
Reduce the chunk size to 50 tokens
Why it's wrong here
Chunk size governs retrieval granularity and context precision, not document recency; 50-token chunks cannot make stale source content current. It tempts because smaller chunks often sharpen semantic matching, and would be the right lever when answers cite irrelevant passages rather than outdated ones.
✗
Increase the model's temperature to 1.0
Why it's wrong here
Temperature controls sampling randomness in token selection, so raising it to 1.0 yields more varied wording, never fresher facts. It tempts because higher values are commonly tuned for creative generation, and would be correct when a chatbot's replies feel repetitive or overly deterministic.
✓
Implement document chunking with metadata such as version or timestamp
Why this is correct
Chunking documents while attaching version or timestamp metadata lets retrieval filter or rank by recency, so superseded content is deprioritised. This directly addresses the freshness constraint by ensuring the index distinguishes current documents from stale ones.
✗
Disable grounding and rely on the model's pre-training data
Why it's wrong here
Disabling grounding discards the retrieval layer entirely, so the model answers from static pre-training weights that cannot reflect recent internal changes — the opposite of improving freshness. It tempts because pre-training data needs no index maintenance, and would suit purely general-knowledge chatbots where no proprietary corpus exists.

Go deeper

Generative AI Leader Practice Question: Using Vertex AI RAG Engine to ground a chatbot in…

Why each option matters

Option-by-option breakdown

Related to this question