Generative AI Leader Practice Question: Techniques to Improve Generative AI Model Output
A retail analytics team uses a Gemini model to answer questions over a large product catalog stored in BigQuery. Answers are sometimes outdated because the model relies on its training data. They want responses to reflect the latest catalog rows and cite the source table. Which technique should they implement?
⚠ Common exam trap
The trap here is thinking that a larger context window or fine-tuning can substitute for live retrieval, when only query-time retrieval guarantees current data and citations.
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
✓
Use retrieval-augmented generation by querying BigQuery and inserting the retrieved rows into the prompt as grounding context.
Retrieval-augmented generation grounds the model in data fetched at request time, which keeps answers current and allows the source table to be cited. Fine-tuning freezes a stale snapshot, stuffing the whole catalog is impractical, and temperature changes only affect randomness. When answers must track a live system of record, retrieval is the appropriate pattern.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Increase the model's context window and paste the entire product catalog into every prompt so the model always has all data.
Why it's wrong here
Even large context windows cannot hold a full enterprise catalog, and stuffing irrelevant rows degrades attention and raises cost and latency. It also provides no clean citation mechanism. This approach does not scale and can reduce answer quality as important details get lost among thousands of unrelated products.
- ✗
Fine-tune the model on a recent export of the product catalog so the weights contain current product data.
Why it's wrong here
Fine-tuning bakes a snapshot of data into the weights, so it becomes stale as soon as the catalog changes and it is expensive to repeat. It also does not provide citations to a source table. For frequently updated inventory, tuning is the wrong tool because it cannot guarantee freshness and gives no traceable provenance for each answer.
- ✗
Enable a higher temperature so the model can combine its training knowledge with newer patterns and produce fresher answers.
Why it's wrong here
Temperature controls randomness, not access to data. Raising it cannot introduce catalog rows the model never saw and will likely increase fabricated details. Freshness requires an external data source at inference time, so this setting change fails to address the stated problem and may worsen hallucination.
- ✓
Use retrieval-augmented generation by querying BigQuery and inserting the retrieved rows into the prompt as grounding context.
Why this is correct
Retrieval-augmented generation fetches current rows at query time and passes them to the model as context, so answers reflect the live catalog. Because the retrieved table and rows can be named in the prompt or returned alongside the answer, citations are possible. This solves both freshness and provenance without retraining.
Go deeper
Related to this question
About these practice questions
This Generative AI Leader question is part of Courseiva's 1,008-question bank — original exam-style content with full explanations and wrong-answer analysis, never real exam questions or exam dumps. Learn why practice questions differ from exam dumps →
JA
Written and reviewed by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
Last reviewed September 2026 · checked against the official Google Cloud exam blueprint
This Generative AI Leader practice question is part of Courseiva's free Google Cloud certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the Generative AI Leader exam.