Monitoring and Logging for Deployed Applications GENERATIVE-AI-ENGINEER-ASSOCIATE Practice Questions (Page 4)

Questions 16–20

1. 16expert · hard

   A company has deployed an agent that uses a mix of LLM calls and deterministic code. They want to monitor the agent's performance, but they are concerned that enabling Agent Monitoring might capture sensitive data in the logs. They need to comply with data privacy regulations. Which approach should they take?

   • AEnable Agent Monitoring but configure redaction rules to mask sensitive data in the logs.
   • BDo not log any data and rely on external monitoring tools.
   • CUse AI Gateway inference tables and manually redact data before storing.
   • DDisable Agent Monitoring and rely on inference logging on the LLM endpoints.
   Reveal Solution Discussion Select an answer first
2. 17application · medium

   A company provides an LLM API to external partners. They want to ensure that no single partner can monopolize the endpoint's capacity. They also want to track how many requests each partner makes to enforce quotas. Which configuration should they implement?

   • AEnable inference logging for all requests.
   • BSet a timeout on the endpoint to limit request duration.
   • CUse Agent Monitoring to observe partner usage.
   • DSet up rate limiting policies per partner and log usage to usage tables.
   Reveal Solution Discussion Select an answer first
4. 18application · medium

   A company has deployed an agent that answers customer questions by calling multiple tools. The support team reports that the agent sometimes gives incorrect answers. The ML team wants to see the chain of thought, tool calls, and final response for each user interaction to diagnose the issue. Which feature should they use?

   • AQuery the agent's training logs
   • BAI Gateway usage tables
   • CAgent Monitoring
   • DInference logging on the underlying LLM endpoint
   Reveal Solution Discussion Select an answer first
5. 19foundation · easy

   What is the primary purpose of rate limiting with AI Gateway?

   • ATo log all requests for auditing purposes
   • BTo automatically scale the endpoint based on demand
   • CTo improve the quality of the model's responses
   • DTo control traffic and ensure fair usage of deployed LLM or agent endpoints
   Reveal Solution Discussion Select an answer first
6. 20foundation · easy

   What is the primary purpose of Agent Monitoring in Databricks?

   • ATo automatically retrain the agent when performance drops
   • BTo manage the infrastructure that hosts the agent
   • CTo observe and evaluate the performance of a deployed agent in real time
   • DTo enforce security policies on the agent's inputs
   Reveal Solution Discussion Select an answer first