AI-103 Practice Test 5 — 25 Questions

Question 1

You are deploying a model in Azure AI Foundry. You need to ensure that the model endpoint is protected by a virtual network. Which configuration should you use?

Accepted Answer

Configure a Private Endpoint for the AI resource. Private endpoints provide secure access to Azure AI resources within a virtual network.

Answer

Enable public internet access with IP firewall rules

Answer

Enable Managed Identity on the endpoint

Answer

Deploy a dedicated Azure Application Gateway

Question 2

You need to extract data from a highly varied set of tax forms where the layout changes frequently. Which approach is most efficient?

Accepted Answer

Use a Neural custom model. A neural custom model is best suited for documents with high visual variability.

Answer

Use a Template model

Answer

Use Read API

Answer

Use General Document model

Question 3

When configuring Azure OpenAI Custom Model Evaluation in Azure AI Foundry, which TWO metrics can be computed using LLM-as-a-judge evaluators? (Choose two.)

Accepted Answer

Coherence. Coherence and relevance are standard LLM-as-a-judge evaluation metrics in Azure AI Foundry.

Answer

CPU utilization percentage

Answer

DNS propagation delay

Answer

Disk input/output operations per second (IOPS)

Question 4

You need to ensure that your agent's response is formatted as valid JSON. How should you approach this?

Accepted Answer

Define the JSON schema in the system prompt instructions.. Using structured outputs or defining the schema in the system prompt is the standard way to enforce output formats in LLMs.

Answer

Set the temperature to 0.5.

Answer

Use an external library to convert text to JSON.

Answer

Ask the user to request JSON.

Question 5

When evaluating an AI agent, what does the 'Coherence' metric measure?

Accepted Answer

The logical flow and clarity of the generated text.. Coherence measures how well the generated text flows and connects logical ideas together.

Answer

The cost of the API call.

Answer

The number of sources cited.

Answer

The speed of the response.

Question 6

Which TWO features in Azure AI Foundry help you manage the lifecycle of your AI agents?

Accepted Answer

Model deployment versions.. Managing lifecycle involves both versioning (deployment) and testing (evaluation).

Answer

The Azure OpenAI Playground history.

Answer

Azure AI Content Safety logs.

Answer

The global search index.

Question 7

You are building an agent in Azure AI Foundry that needs to retrieve information from a company SharePoint site. Which component should you implement to ground the agent?

Accepted Answer

Azure AI Search with a SharePoint connector. Azure AI Search with a SharePoint indexer is the standard way to ground agents in external document stores.

Answer

Blob Storage integration

Answer

Direct REST API call to SharePoint

Answer

Microsoft Graph Data Connect directly to the model

Question 8

When implementing Retrieval-Augmented Generation (RAG) using Azure AI Search and Azure OpenAI, which THREE strategies help optimize chunk retrieval relevance and reduce noise? (Choose three)

Accepted Answer

Configure hybrid search combining vector embeddings and keyword full-text search.. Hybrid search (combining vectors and keywords), semantic ranking, and appropriate chunk sizing are proven strategies for optimizing RAG relevance.

Answer

Store all corporate documents in a single unindexed raw text blob without metadata fields.

Answer

Set the embedding model dimensions to the lowest possible setting to maximize retrieval speed.

Question 9

You are implementing prompt engineering for a customer support agent. To minimize hallucinations when the model lacks information, which Prompt Flow technique is most effective?

Accepted Answer

Instructional grounding to only answer from context. Prompt engineering for RAG requires explicit instructions to use provided context and say 'I don't know' if missing.

Answer

Chain-of-thought prompting

Answer

Increasing the Top_P parameter

Answer

Few-shot prompting with varied examples

Question 10

You are troubleshooting a Prompt Flow and suspect the issue lies in the input processing. Which THREE tools or techniques can you use?

Accepted Answer

Setting breakpoints in the visual designer.. Debugging in Prompt Flow involves visual inspection of node inputs/outputs, tracing, and log analysis.

Answer

Deleting and recreating the entire resource group.

Answer

Increasing the model temperature to see if the error persists.

Question 11

You are building a RAG solution and observe that the model retrieves the correct documents but fails to use the information accurately. What is the most effective approach to improve performance?

Accepted Answer

Refine the system prompt instructions. Iterative prompt engineering and improving the system message instructions are key to improving RAG performance.

Answer

Increase the vector index size

Answer

Use a smaller base model

Answer

Disable streaming response

Question 12

Your company requires strict data privacy. Where should you host your Document Intelligence resources to ensure data remains in a specific region?

Accepted Answer

Regional Azure resource. Choosing the appropriate Azure region during resource creation ensures data residency compliance.

Answer

Any available region

Answer

Global service endpoint

Answer

Public multi-region endpoint

Question 13

Which TWO of the following are prerequisites for training a custom extraction model?

Accepted Answer

A set of labeled data in JSON format. You need a source of files in Azure Blob Storage and the files must be associated with valid labels (via .labels.json).

Answer

An Azure SQL database

Answer

A pre-existing template model

Answer

A pre-installed local labeling agent

Question 14

You are building an agent in Azure AI Foundry that requires access to a private internal database. Which component should you configure to ensure the agent retrieves data securely without exposing the database endpoint publicly?

Accepted Answer

Azure Private Link and Managed Virtual Network. Managed Virtual Networks and Private Endpoints in Azure AI Foundry ensure that traffic between your agent and data sources remains within the Microsoft backbone network.

Answer

Public IP allowlisting

Answer

Entra ID Application Proxy

Answer

Shared Access Signatures (SAS)

Question 15

Which of the following is a core principle of Responsible AI in Azure AI Foundry?

Accepted Answer

Transparency and explainability.. Transparency is a core principle, often implemented by providing sources and citations in AI responses.

Answer

Bypassing all safety filters.

Answer

Maximizing cost savings at any cost.

Answer

Forcing the model to agree with the user.

Question 16

You are building an agent in Azure AI Foundry using a prompt flow. You need to ensure that the agent only uses company-approved documentation for its responses. Which feature should you implement to meet this requirement?

Accepted Answer

Configure Grounding with data. Grounding with data is the primary mechanism in Azure AI Foundry to limit an agent's knowledge to specific data sources.

Answer

Enable prompt tuning

Answer

Implement system message constraints

Answer

Deploy a fine-tuned model

Question 17

Which TWO actions should you take to ensure your agentic solution is 'Responsible'?

Accepted Answer

Enable Content Safety filters for all inputs and outputs.. Responsible AI involves both technical guardrails (Content Safety) and transparency (disclosures/citations).

Answer

Disable all logging to protect user privacy.

Answer

Set the model temperature to 1.0 to increase diversity.

Answer

Allow the agent to access the public internet without restrictions.

Question 18

Which TWO resources are required to set up an Azure AI Foundry project?

Accepted Answer

An Azure AI Services or Azure OpenAI resource.. A project requires a backing resource (Azure AI Services or Azure OpenAI) and a storage mechanism (Azure Storage) to function.

Answer

An Azure Data Factory instance.

Answer

A virtual machine.

Answer

A dedicated Kubernetes cluster.

Question 19

Which component in Azure AI Foundry allows you to visualize and debug the execution path of a model call, including input/output mapping?

Accepted Answer

Prompt Flow. Prompt Flow provides the visual designer and debugger for workflows.

Answer

Compute Instance

Answer

Model Benchmarks

Answer

Deployment Endpoint

Question 20

You are configuring an agent that must call a REST API to get current stock prices. Where must you define this integration to ensure the agent can reliably call the tool?

Accepted Answer

In the Agent Tool Definition.. Tool definitions are required for models to understand how to interact with external functions.

Answer

In the Azure OpenAI deployment configuration.

Answer

In the Search index field mapping.

Answer

In the System Prompt.