Which Azure AI Speech feature allows for "Speaker Diarization" (identifying who spoke when)?
check_circle
Correct AnswerB
Batch Transcription includes a diarization feature that labels different speakers in an audio file.
Question 42
You want to implement "Token-based Rate Limiting" for your agent to prevent a single user from exhausting your Azure OpenAI quota. Where should this be implemented?
check_circle
Correct AnswerB
API Management (APIM) is the standard tool for implementing rate limiting, throttling, and usage policies for Azure APIs.
Question 43
Which Azure service is used to host the "Custom API" that your agent calls to check real-time inventory levels?
check_circle
Correct AnswerC
Azure Functions (or App Service) provides the serverless environment to run the code that your agent's Function Tool triggers.
Question 44
You are using "Prompt Flow" and want to implement a logic gate that only calls the LLM if the user's query is related to "Finance". Which node should you use?
check_circle
Correct AnswerB
Python nodes allow for custom code-based routing and filtering before data reaches the next stage of the flow.
Question 45
You are deploying an AI agent and want to minimize costs for a testing environment with highly intermittent traffic. Which pricing model for Azure OpenAI models is most appropriate?
check_circle
Correct AnswerB
Serverless API charges per token processed, making it the most cost-effective choice for low-volume or unpredictable traffic.
Question 46
You need to identify if a model's output contains a "Direct Copy" of copyrighted text. Which tool in the Azure AI portfolio does this?
check_circle
Correct AnswerC
This feature in Azure AI Content Safety checks generated text against a database of known proprietary and copyrighted content.
Question 47
Which Azure OpenAI model supports "Global Deployment" to automatically route requests to regions with available quota?
check_circle
Correct AnswerD
Global-Standard deployments (especially GPT-4o) allow for high-throughput routing across Azure's global infrastructure.
Question 48
Which metric in AI Foundry Evaluation measures the time taken from the user's initial request to the first token generated by the model?
check_circle
Correct AnswerC
TTFT is a critical performance metric for user experience, measuring the responsiveness of the generative AI system.
Question 49
You are using "Prompt Flow" and want to implement a logic gate that only calls the LLM if the user's query is related to "Finance". Which node should you use?
check_circle
Correct AnswerB
Python nodes allow for custom code-based routing and filtering before data reaches the next stage of the flow.
Question 50
You want to give your agent the ability to read and describe a complex architectural diagram in an image. Which model is best?
check_circle
Correct AnswerD
GPT-4o’s vision capabilities allow it to perform high-level reasoning and data extraction from visual inputs like diagrams.
