A startup uses the AI Gateway to route requests to multiple LLM providers. They want to log the full request and response payloads for all traffic to a central location for auditing and debugging. Which configuration should they use?

Use the provider's native logging console.

Enable AI Gateway usage tables to track token consumption.

Configure AI Gateway inference tables to capture request and response data.

Set up a separate logging service to capture the data.

Which of the following is a key feature of Agent Monitoring for deployed agents?

Integration with external CI/CD pipelines for deployment

Automatic model versioning and rollback

Automated prompt engineering for better responses

Real-time dashboards showing agent performance metrics

A team has enabled inference logging for a live LLM endpoint. They want to identify requests that resulted in a specific error message in the response. Which query approach should they use on the inference table?

Query the model's training logs for error patterns.

Use the usage table to find requests with high token counts.

Filter the request column for the error message text.

Filter the response column for the error message text.

A company uses the AI Gateway to route requests to a fine-tuned model. They want to monitor the quality of responses by comparing the model's output to a golden set of expected answers. They also need to track the cost of these evaluation requests separately from production traffic. Which combination of features should they use?

Enable inference tables for the evaluation traffic and tag the requests with a 'eval' label, then use usage tables to track costs.

Set up a separate endpoint for evaluation and use only usage tables.

Use Agent Monitoring to evaluate the model's responses.

Enable rate limiting to control evaluation traffic.

How can you query an inference table to analyze the behavior of a live LLM endpoint?

Using the MLflow tracking API to retrieve metrics

Using SQL queries in a Databricks notebook or SQL warehouse

Using the model serving UI to view individual requests

Using the Databricks CLI to export the table to a CSV file

DatabricksCertified Generative AI Engineer Associate

Domain 6Objective 2

Monitoring and Logging for Deployed Applications GENERATIVE-AI-ENGINEER-ASSOCIATE Practice Questions (Page 5)

Part of the Section 6: Evaluation and Monitoring domain, which makes up ~17% of our current practice bank. Databricks does not publish an official question count, but from its 90-minute exam (~35–60 total, ~6–10 in this domain), expect 1–2 from this objective — we provide 25 practice questions to prepare you well beyond it. (estimate)

25questions here

5free pages

6concepts

Questions 21–25

21application · medium
A startup uses the AI Gateway to route requests to multiple LLM providers. They want to log the full request and response payloads for all traffic to a central location for auditing and debugging. Which configuration should they use?
Select an answer first
22foundation · easy
Which of the following is a key feature of Agent Monitoring for deployed agents?
Select an answer first
23application · medium
A team has enabled inference logging for a live LLM endpoint. They want to identify requests that resulted in a specific error message in the response. Which query approach should they use on the inference table?
Select an answer first
24application · medium
A company uses the AI Gateway to route requests to a fine-tuned model. They want to monitor the quality of responses by comparing the model's output to a golden set of expected answers. They also need to track the cost of these evaluation requests separately from production traffic. Which combination of features should they use?
Select an answer first
25foundation · easy
How can you query an inference table to analyze the behavior of a live LLM endpoint?
Select an answer first

Finished these 5 questions?

Review the revealed explanations, or continue through the curriculum.

No more pagesBack to GENERATIVE-AI-ENGINEER-ASSOCIATE

Free Basic Practice is a study aid with revealable answers — not a scored exam. Examers.io is independent and not affiliated with or endorsed by Databricks. “GENERATIVE-AI-ENGINEER-ASSOCIATE” is a trademark of its owner, used for identification only.