AI-102 Implement generative AI solutions Practice Test 5 — 15 Questions

Question 1

A development team is using the Azure OpenAI Assistants API to build an agent that answers questions over a large set of internal technical manuals. The agent must reference the manuals during a conversation and must be able to read and write files stored in an Azure Storage account. The team wants the least-effort integration that keeps the manual content managed by Azure rather than uploaded per session. Which tool should the team attach to the assistant?

Choose an answer to begin — your selection is scored in the full session.

Accepted Answer

The function calling tool with a function that accesses Azure Storage. The distinguishing requirement is that the agent must read and write files in Azure Storage, which is an action rather than a retrieval task. Function calling is the Assistants API mechanism for letting the model invoke developer-defined code, so exposing storage operations as functions satisfies that need. Retrieval-oriented tools such as file search or Azure AI Search ground answers but cannot perform the storage writes, and code interpreter runs isolated sandbox code without a storage connector.

Answer

The code interpreter tool

Answer

The Azure AI Search tool connected to an index built over the manuals

Answer

The file search tool with the manuals uploaded as assistant files

Question 2

You are deploying a GPT-4o model in Azure OpenAI Service for an internal knowledge assistant. The assistant must return answers only when the retrieved source documents have a high similarity score, and it must explicitly state when no relevant documents are found. You need to configure the Azure OpenAI On Your Data request to enforce this behavior. Which setting should you configure?

Accepted Answer

Set the 'score_threshold' parameter and the 'no_documents_message' parameter in the data source configuration.. The score_threshold parameter filters retrieved chunks by relevance, and no_documents_message provides a deterministic response when nothing passes the filter. Using both satisfies the dual requirement of high-similarity-only answers and an explicit notice when no relevant documents exist.

Answer

Set the 'strictness' parameter to a high value and set 'in_scope' to true.

Answer

Set 'top_p' to 0.1 and add a system message instructing the model to say it does not know.

Answer

Set 'filter' to exclude documents below a score and configure 'retrieval_queries'.

Question 3

You are deploying an Azure OpenAI chat solution for a bank's internal knowledge assistant. Compliance requires that every generated answer be grounded only in the approved policy index, and that the model refuse to answer when no approved content is retrieved. You need to enforce this with the least development effort. What should you configure on the model deployment?

Accepted Answer

Add an 'On Your Data' data source to the deployment and enable the strictness setting so the model only uses retrieved content.. Grounding an Azure OpenAI chat deployment in an approved index is done by attaching an 'On Your Data' data source and enabling strictness, which limits responses to retrieved content and produces a refusal when nothing relevant is found. This meets the compliance goal with configuration instead of custom orchestration code.

Answer

Set the deployment's content filter to block the 'Violence' category at high severity.

Answer

Create a fine-tuned model from the policy documents and deploy it as the chat model.

Answer

Increase the deployment's tokens-per-minute quota so retrieval results always fit in the context window.

Question 4

You deploy a GPT-4o model in Azure OpenAI Service for an internal assistant. Compliance requires that prompts and completions are never used to improve Microsoft or OpenAI models and that data stays within your Azure region. You also need the ability to run abuse-monitoring opt-out for a regulated workload. Which combination of configurations satisfies these requirements?

Accepted Answer

An Azure OpenAI resource deployed in the required region with the Limited Access approval and the abuse-monitoring opt-out enabled for the subscription.. Azure OpenAI does not train foundation models on customer prompts or completions, and data residency follows the resource region. The remaining concern is abuse monitoring, which by default can retain data for human review. Applying for and enabling the abuse-monitoring opt-out, along with deploying the resource in the required region, satisfies the no-training, residency, and opt-out requirements simultaneously.

Answer

An Azure OpenAI resource with diagnostic settings streaming logs to a Log Analytics workspace in the same region.

Answer

An Azure OpenAI resource with customer-managed keys stored in Azure Key Vault and a private endpoint for network isolation.

Answer

A standard Azure OpenAI resource with the default abuse monitoring setting and a content filter policy set to medium severity.

Question 5

You are building a generative AI application using Azure OpenAI Service. The application must generate responses that include citations from a set of internal documents stored in Azure Cosmos DB. You want to use the 'On Your Data' feature. Which data source type should you configure?

Accepted Answer

Azure Cosmos DB. The On Your Data feature in Azure OpenAI Service supports Azure Cosmos DB as a data source. Configuring the data source type to Azure Cosmos DB allows the model to retrieve relevant documents directly from the database and include citations in its responses. This meets the requirement without needing to index the documents into another service.

Answer

Azure Cognitive Search

Answer

Azure SQL Database

Answer

Azure Blob Storage

Question 6

You are deploying a generative AI chatbot using Azure OpenAI Service. The chatbot must not generate responses that include profanity or hate speech. You need to configure content filtering to block harmful content. What should you do?

Accepted Answer

Enable the default content filters in the Azure OpenAI deployment and adjust the severity thresholds as needed.. The correct method is to enable and configure Azure OpenAI's built-in content filters. These filters automatically detect and block harmful categories like profanity and hate speech, and you can adjust severity thresholds to match your policy. They provide a robust, integrated safeguard without custom development, ensuring the chatbot complies with content safety requirements.

Answer

Use prompt engineering to instruct the model to avoid profanity and hate speech.

Answer

Set the model temperature to 0 to reduce the likelihood of harmful content.

Answer

Implement a custom blocklist in Azure AI Content Safety and apply it to the deployment.

Question 7

You operate an Azure OpenAI chat app that calls a deployment through the Azure OpenAI SDK. A new compliance rule requires that each request be tagged with the calling user's department and that this tag appear in Azure Monitor logs for auditing. You must avoid changing application code and avoid sending the department as part of the chat message. What should you implement?

Accepted Answer

Place an Azure API Management gateway in front of the Azure OpenAI deployment and add the department as a custom dimension using a policy.. Tagging requests without touching application code is best achieved by inserting Azure API Management in front of the Azure OpenAI deployment and using a policy to attach the department value, which then flows into logs and metrics. This keeps the chat payload clean and provides a structured, auditable field.

Answer

Enable diagnostic settings on the Azure OpenAI resource and rely on the default request logs.

Answer

Create a custom header or metadata field in the API request and configure logging to capture it.

Answer

Add the department to the system message of each request before sending it.

Question 8

You are deploying a generative AI chat application on Azure OpenAI Service. The application must log every prompt and completion to an Azure Monitor Log Analytics workspace for auditing, without modifying application code to call logging APIs. Which feature should you enable?

Accepted Answer

Diagnostic settings on the Azure OpenAI resource that send logs and metrics to the Log Analytics workspace.. Resource-level diagnostic settings stream Azure OpenAI request and response logs to Log Analytics without any client code changes. Enabling the audit log category captures prompts and completions, which provides the required centralized audit trail while leaving the application untouched.

Answer

Content filtering configuration set to log blocked categories.

Answer

Application Insights SDK instrumentation added to the client application.

Answer

Azure Monitor alerts configured on token usage metrics for the deployment.

Question 9

A company wants to generate personalized product descriptions for its e-commerce site using Azure OpenAI. They need to ensure the model's output adheres to brand guidelines and does not generate prohibited content. Which approach should they use?

Accepted Answer

Use a system message with brand guidelines and apply content filtering.. Using a system message allows you to embed brand guidelines directly into the conversation context, instructing the model on tone, style, and prohibited content. Azure OpenAI's content filtering provides an additional safety layer by automatically detecting and blocking harmful or policy-violating outputs, ensuring compliance with both brand and regulatory requirements.

Answer

Use prompt engineering with negative prompts and ignore content filtering.

Answer

Provide few-shot examples in the user message and rely on the model's training.

Answer

Fine-tune the model with brand guidelines and disable content filtering for performance.

Question 10

A healthcare startup is developing a chatbot that uses Azure OpenAI to answer patient questions. They need to ensure that the chatbot only uses information from their verified medical database and does not generate unsupported medical advice. What is the best approach?

Accepted Answer

Use Azure AI Search with vector search to retrieve relevant documents and pass them as context.. It uses Azure AI Search with vector search to retrieve only relevant, verified documents from the medical database and passes them as context to the Azure OpenAI model. This grounds the model's responses in authoritative data, preventing it from generating unsupported medical advice. The retrieval-augmented generation (RAG) pattern ensures the chatbot answers are based on the provided context rather than the model's internal knowledge.

Answer

Fine-tune a model on the medical database and deploy it.

Answer

Embed the entire medical database in the system message.

Answer

Rely on Azure OpenAI's content filtering to block unsupported advice.

Question 11

A financial services firm wants to use Azure OpenAI to generate investment advice summaries. They must ensure that the model does not produce any advice that could be interpreted as personalized financial advice. What is the most effective strategy?

Accepted Answer

Use a system message that instructs the model to avoid personalized advice and apply strict content filtering.. Azure OpenAI's system messages allow you to set the model's behavior and constraints at the conversation level, which is the most direct and effective way to enforce a policy like avoiding personalized financial advice. Combined with Azure's content filtering (which can block harmful or restricted content), this approach provides both instruction-based and filter-based guardrails without requiring model retraining or relying solely on example-based prompting.

Answer

Set temperature to 0 and top_p to 0 to make outputs deterministic.

Answer

Provide few-shot examples of disclaimers in the prompt.

Answer

Fine-tune the model on a dataset of generic financial summaries.

Question 12

A developer wants to use Azure OpenAI to generate text from a prompt. Which parameter controls the diversity of the generated output?

Accepted Answer

temperature. Temperature is the parameter that directly controls the randomness or diversity of the generated output by scaling the logits before applying the softmax function. A higher temperature (e.g., 1.0) increases the probability of less likely tokens, producing more creative and varied responses, while a lower temperature (e.g., 0.1) makes the output more deterministic and focused.

Answer

presence_penalty

Answer

frequency_penalty

Answer

max_tokens

Question 13

A company is using Azure OpenAI to generate customer support responses. They want to ensure the model does not use any personally identifiable information (PII) in its outputs. What should they implement?

Accepted Answer

Use Azure AI Content Safety to filter PII from the output.. Azure AI Content Safety provides built-in PII detection and redaction capabilities that can automatically scan and filter sensitive information from model outputs. This is the most reliable approach because it operates as a post-processing filter, catching PII that the model might generate despite instructions. Fine-tuning, prompt engineering, and system messages are all fallible because they rely on the model's compliance rather than enforced filtering.

Answer

Fine-tune the model on anonymized data.

Answer

Use prompt engineering to instruct the model to redact PII.

Answer

Use a system message instructing the model to avoid PII.

Question 14

Which TWO actions should you take to reduce the cost of using Azure OpenAI for a chatbot that handles high traffic?

Accepted Answer

Use the batch API for non-real-time requests.. Option B is correct because the Azure OpenAI Batch API processes asynchronous, non-real-time workloads at a 50% discount compared to standard global pricing, so routing any chatbot requests that don't need immediate responses through batch jobs directly lowers per-token cost. Option C is correct because caching responses to frequently asked questions (for example with Azure Cache for Redis or a semantic cache) means repeated identical or similar prompts are served without calling the model at all, eliminating token charges for that traffic and reducing overall request volume. Option A is not correct because increasing max_tokens raises the maximum output length and therefore can increase, not decrease, token consumption and cost. Option D is not correct because temperature controls randomness in sampling, not the number of tokens generated, so setting it to 0 does not reduce token usage or cost. Option E is not correct because fine-tuning does not shorten the prompt by itself and adds training and hosting costs; prompt length is reduced through techniques like prompt compression or shorter system messages, not fine-tuning alone.

Answer

Increase max_tokens to reduce the number of requests.

Answer

Lower the temperature to 0 to reduce token usage.

Answer

Fine-tune the model to reduce prompt length.

Question 15

Which THREE factors should you consider when selecting a model for a generative AI solution on Azure?

Accepted Answer

Cost per token and deployment options.. Option A is correct because cost per token and deployment options (such as pay-as-you-go versus provisioned throughput) directly affect the total cost and scalability of a generative AI solution on Azure. Option B is correct because the model's capability and modality determine whether it can handle the required task, such as text generation, code completion, or image creation. Option C is correct because latency and throughput requirements dictate whether the chosen model and deployment type can meet the application's performance and concurrency needs. Option D is not a primary selection factor because the number of transformer layers is an internal architectural detail that influences capability but is not a decision criterion by itself. Option E is not a primary selection factor because training data source and licensing are legal and compliance considerations, not core factors for selecting a model for a generative AI solution on Azure.

Answer

Number of transformer layers in the model.

Answer

Training data source and licensing.