A data engineer wants to enable inference logging for a deployed RAG application on Databricks. What is the primary purpose of enabling inference logging?

To capture request and response data for assessing the performance of the deployed RAG application

To automatically retrain the model when performance degrades

To store the model weights and artifacts for version control

To enforce rate limits on the endpoint to prevent overuse

An organization exposes an LLM endpoint through the AI Gateway. To prevent a single user from consuming excessive resources, they need to limit the number of requests per user per minute. Which configuration should they implement?

Configure a rate limiting policy in the AI Gateway based on user identity.

Enable inference logging to track user activity.

Use usage tables to monitor and manually block users.

Set a timeout on the serving endpoint.

A team has a RAG application with inference logging enabled. They want to analyze the quality of answers by comparing the retrieved context to the final response. However, the inference table only stores the request and response, not the retrieved chunks. The team wants to capture the retrieved chunks without changing the application's external behavior. Which approach should they take?

Query the usage table to infer the retrieved chunks from token counts.

Use Agent Monitoring to capture the retrieval step.

Enable a separate logging feature in the vector database.

Modify the application to include the retrieved chunks in the request payload sent to the endpoint.

A company has deployed an agent that uses a custom tool to fetch real-time inventory data. The agent sometimes fails to call the tool correctly. The team wants to see the exact tool call arguments and the tool's response for each failed interaction. Which monitoring approach provides this detail?

Use Agent Monitoring to view the full trace of the agent's execution.

Add logging statements to the tool's code.

Enable inference logging on the underlying LLM endpoint.

Query the AI Gateway usage table for tool call metadata.

DatabricksCertified Generative AI Engineer Associate

Domain 6Objective 2

Monitoring and Logging for Deployed Applications GENERATIVE-AI-ENGINEER-ASSOCIATE Practice Questions (Page 2)

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 6–10

6foundation · easy
A data engineer wants to enable inference logging for a deployed RAG application on Databricks. What is the primary purpose of enabling inference logging?
Select an answer first
7application · medium
An organization exposes an LLM endpoint through the AI Gateway. To prevent a single user from consuming excessive resources, they need to limit the number of requests per user per minute. Which configuration should they implement?
Select an answer first
8expert · hard
A team has a RAG application with inference logging enabled. They want to analyze the quality of answers by comparing the retrieved context to the final response. However, the inference table only stores the request and response, not the retrieved chunks. The team wants to capture the retrieved chunks without changing the application's external behavior. Which approach should they take?
Select an answer first
9application · medium
A company has deployed an agent that uses a custom tool to fetch real-time inventory data. The agent sometimes fails to call the tool correctly. The team wants to see the exact tool call arguments and the tool's response for each failed interaction. Which monitoring approach provides this detail?
Select an answer first
10application · medium
A finance team needs to allocate LLM costs to different departments. They have an AI Gateway that routes requests for multiple internal applications. Which data source should they use to calculate the cost per department?
Select an answer first

Finished these 5 questions?

Review the revealed explanations, or continue through the curriculum.

Next 5 questions →Back 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.

Monitoring and Logging for Deployed Applications GENERATIVE-AI-ENGINEER-ASSOCIATE Practice Questions (Page 2)

Questions 6–10

A data engineer wants to enable inference logging for a deployed RAG application on Databricks. What is the primary purpose of enabling inference logging?

An organization exposes an LLM endpoint through the AI Gateway. To prevent a single user from consuming excessive resources, they need to limit the number of requests per user per minute. Which configuration should they implement?

A company has deployed an agent that uses a custom tool to fetch real-time inventory data. The agent sometimes fails to call the tool correctly. The team wants to see the exact tool call arguments and the tool's response for each failed interaction. Which monitoring approach provides this detail?

A finance team needs to allocate LLM costs to different departments. They have an AI Gateway that routes requests for multiple internal applications. Which data source should they use to calculate the cost per department?

Keep studying — Section 6: Evaluation and Monitoring