Free Microsoft AI-103 Practice Questions

Question 1

You are planning a Microsoft Foundry project named Project1 that will contain multiple agents. Each agent will access the same Azure AI Search resource.

You need to recommend a solution to centrally manage the Azure AI Search credentials within Project1. The solution must be implemented across all the agents.

What should you recommend?

1. Enable role-based access control (RBAC) for the Azure AI Search resource.
2. Disable key-based access control on the Azure AI Search resource.
3. Add a connection to the Azure AI Search resource.
4. Create a managed private endpoint that connects to the Azure AI Search resource.

Question 2

You have a Microsoft Foundry project that contains an agent. The agent uses Azure Speech in Foundry Tools.

You fine-tune a baseline speech to text model for the en-us locale and publish the model.

The agent calls the Speech to text REST API and returns an error message indicating that the project ID is invalid.

You need to set the project property to the correct ID.

To what should you set the project property?

1. the project URL
2. the custom speech project ID
3. the project ID
4. the custom speech endpoint URL

Question 3

You have an Azure Speech in Foundry Tools resource that hosts a custom speech to text model deployed to a custom endpoint. An agent uses the endpoint to perform real-time speech recognition.

You are approaching the expiration date of the custom speech to text model.

What is the expected behavior when the model expires?

1. Speech recognition requests will return a 4xx error until a new custom model is deployed.
2. Speech recognition requests will continue to use the expired custom model until the model is removed manually.
3. Speech recognition requests will fall back to the most recent base model for the same locale.
4. The custom model will be deleted automatically when the model expires.

Question 4

You have a Microsoft Foundry project that contains a model deployment.

You have an application that calls the deployment by using the Azure OpenAI v1 API and DefaultAzureCredential.

The developers at your company receive HTTP 403 errors when they send inference requests, even after running az login.

You need to ensure that the developers can perform model inference. The solution must follow the principle of least privilege.

Which role-based access control (RBAC) role should you assign to the developers?

1. Cognitive Services User
2. Cognitive Services OpenAI User
3. Contributor
4. Cognitive Services Data Reader

Question 5

You have a Microsoft Foundry project named Project1 that contains an agent. The agent uses an OpenAPI 3.0 specification to call an external weather service.

The weather service requires a key to be passed in an HTTP header. The key value is stored as a connection in Project1.

You need to ensure that the key value from the connection is included automatically whenever the OpenAPI tool is invoked.

What should you configure in the OpenAPI specification?

1. a header parameter defined for each operation
2. an Azure Key Vault connection
3. an API key security scheme
4. a Bearer token security scheme

Question 6

You have a Microsoft Foundry project that serves a high-volume chat app.

Most requests are simple FAQs, but some require advanced reasoning.

You need to reduce costs and latency for common queries, without degrading the quality of the responses to complex questions.

What should you do?

1. Route all the requests to a smaller model.
2. Use a model cascade that routes the requests to different models.
3. Increase the value of the max_tokens parameter for all the requests.
4. Route all the requests to the most capable model.

Question 7

You have a Microsoft Foundry project that contains a high-traffic agent.

After a recent update, operational costs increase significantly.

Monitoring confirms that the volume of user traffic to the agent remains unchanged.

You suspect that changes to the request or response characteristics are causing the increase. You need to identify whether the additional costs are driven by the model input size, the model output size, or expanded tool usage.

Which observability capability should you use?

1. latency
2. evaluation metrics
3. run success rate
4. token usage

Question 8

You have a Microsoft Foundry project named Project1 that contains the following:

• An OpenAPI tool that calls an external API

• A project connection named Connection1 that stores the API key of the external API

When an agent calls the OpenAPI tool, the API returns a 401 unauthorized error, and traces show that the API key header is NOT being sent.

You need to ensure that the OpenAPI tool automatically includes the API key from Connection1 on all requests.

What should you do?

1. Enable identity passthrough so that the tool uses the Microsoft Entra token of the caller.
2. Add the API key header manually to the OpenAPI specification.
3. Configure the tool to use the default connection of Project1.
4. Connect the tool to Connection1.

Question 9

You have a Microsoft Foundry project that contains a customer support agent. The agent calls an internal knowledge API tool before generating responses.

Users report the following issues:

Some requests take more than 15 seconds to complete.

Some responses are incorrect, even when the knowledge API returns the expected data.

You need to inspect individual agent runs to view the ordered sequence of large language model (LLM) calls, tool invocations, and timing information.

Which observability capability should you use?

1. token usage
2. monitoring
3. safety metrics
4. tracing

Question 10

You have a chat app in a Microsoft Foundry project and an Azure AI Search vectorized index.

You need to connect to the index to meet the following requirements:

Complex questions must retrieve information from multiple chunks.

Multi-turn conversations must influence retrieval planning.

Retrievals must run in parallel to reduce latency.

Which retrieval approach should you use?

1. iterative retrieval
2. agentic Retrieval Augmented Generation (RAG)
3. chain of thought
4. classic Retrieval Augmented Generation (RAG)

Question 11

You have a Microsoft Foundry project that contains an agent. The agent generates summaries from retrieved policy documents.

You need to improve response completeness. The solution must be implemented in the logic of the application code before responses are returned.

What should you do?

1. Add a retry evaluation before the responses are returned.
2. Decrease the value of the max_tokens parameter.
3. Switch to Retrieval Augmented Generation (RAG).
4. Replace the model with a smaller deployment.

Question 12

You have a Microsoft Foundry project that contains an agent and uses a GitHub repository. The repository contains a YAML file named File1 that defines the evaluation settings of the agent.

You need to create a GitHub Actions workflow that runs the evaluation defined in File1 when a pull request (PR) is opened.

How should you configure the workflow?

1. Set project-endpoint to the endpoint of the project.
2. Set evaluation-config to the path of the YAML file.
3. Set model-deployment-name to the deployed model.
4. Set tenant-id to the Microsoft Entra tenant ID

Question 13

You have a Microsoft Foundry project that contains an agent.

You need to process mixed-format documents that contain scanned text, tables, and multicolumn layouts. The extracted content must preserve the document structure and be converted into the Markdown format for downstream reasoning.

What should you configure first?

1. an Azure Language in Foundry Tools text analysis model deployment
2. a generative chat completion request
3. an Azure OpenAI Responses API call that uses a multimodal model
4. an Azure Content Understanding in Foundry Tools analyzer

Question 14

You have a Microsoft Foundry project that contains an agent.

The knowledge source for the agent is a set of scanned PDF troubleshooting guides stored in Azure Blob Storage. The guide pages contain two-column layouts and tables.

You use Azure Content Understanding in Foundry Tools to process the PDFs.

You plan to ingest the processed content into an index for Retrieval Augmented Generation (RAG) and store extracted fields for downstream automation.

Stakeholders must be able to verify where each extracted field value came from in the original PDF and route low-reliability extractions for manual review.

You need to ensure that the Content Understanding document analyzer output includes a per-field confidence score and source grounding to locations within the source document.

What should you do?

1. Set enableSegment to true.
2. Provide labeled samples.
3. Enable estimateFieldSourceAndConfidence.
4. Configure the analyzer to use generative extraction for all fields.

Question 15

You have a Microsoft Foundry project that contains an agent.

The agent uses Azure AI Search for Retrieval Augmented Generation (RAG).

You plan to ingest and index PDF product manuals.

You need to build a solution that supports semantic similarity matching. The solution must ensure that the agent retrieves relevant data when user questions use different wording than the product manuals.

Which indexing approach should you use?

1. vector search
2. semantic ranking
3. suggesters
4. analyzers

Question 16

You are developing prompts for a Microsoft Foundry project that classifies incoming support tickets by category.

You need to improve accuracy by showing the model how correct classifications look, without retraining the model or storing knowledge permanently.

Which prompt engineering approach should you use?

1. Retrieval Augmented Generation (RAG)
2. zero-shot learning
3. chain of thought
4. few-shot learning

Question 17

You are building a speech processing solution in Microsoft Foundry for a customer support platform.

The platform will transcribe live phone calls, so that supervisors at your company can view call transcripts and detect issues while the calls are in progress. The call audio will arrive as a continuous stream from the telephony system.

You need to ensure that the call transcripts appear within only a few seconds of the audio stream.

What should you do?

1. Use text to speech by using a custom neural voice.
2. Use speech translation to generate the transcripts into multiple languages.
3. Run a batch transcription job on recorded audio files.
4. Use real-time speech to text to process streaming audio input.

Question 18

You have an application that processes scanned PDF invoices. The invoices have varied layouts and include multipage tables.

You have a pipeline that uses optical character recognition (OCR) and extracts totals and invoice numbers. The results are often incorrect because the document structure is ignored.

You need to implement a solution that provides OCR, layout analysis, and template-generalizing field extraction. The solution must NOT require training a custom model. The solution must minimize administrative effort.

What should you include in the solution?

1. Azure Language in Foundry Tools
2. Azure Content Understanding in Foundry Tools
3. an Azure Machine Learning model

Question 19

You have a Microsoft Foundry project that contains an agent. The agent uses two tools to perform the following actions:

• Use Azure AI Search to retrieve answers from a private product documentation index.

• Use the web search tool to retrieve public information on the internet.

You need to ensure that for a specific run, the agent deterministically retrieves information only from the internet.

To what should you set tool_choice?

1. {“type”: “bing_grounding”}
2. {“type”: “azure_ai_search”}
3. “auto”
4. “required”

Question 20

You have a web app named App1 that processes user prompts by integrating with a Microsoft Foundry project named Project1. App1 performs the following actions:

• Sends prompts directly to a model by using the Azure OpenAI Responses API

• Invokes the Azure AI Content Safety tool by using a Foundry connection within the same request

You need to configure end-to-end visibility into each step of the request workflow.

What should you do?

1. Enable logging by using the client SDK for Content Safety.
2. Enable logging by using Foundry Local.
3. Enable application tracing in Project1.
4. Route requests through the Azure OpenAI endpoint.

Question 21

You have a web app named App1 that sends requests to a multimodal chat model deployment in a Microsoft Foundry project.

User messages can contain both text and images.

Currently, App1 includes image URLs as plain text inside the message content so the model cannot recognize them as images.

Traces show that the requests contain a single text message instead of a multimodal content array.

You need to send the message as a structured array that includes both the text portion and the image reference to ensure that the model can process the image correctly.

What should you do?

1. Set the user message content array to include items that have type: text and type: image_url.
2. Encode the image to base64 and include the encoded data inside the content string of the user message.
3. Add the image URL to the request metadata section, so the model can resolve the processing issue automatically.
4. Place the image URL inside the System Message and set type to image_url so the model loads the image at initialization.

Question 22

You are deploying a support agent that enables users to upload photos.

You need to automatically classify uploaded images for harmful content. The solution must block content based on severity levels.

What should you do?

1. Apply keyword scanning to optical character recognition (OCR) output by using Azure Vision in Foundry Tools.
2. Enable prompt shields.
3. Use blocklists.
4. Implement image moderation.

Question 23

You have a Microsoft Foundry project that contains an agent.

The agent ingests scanned PDF vendor invoices that contain tables and embedded QR codes.

The agent must preserve the PDF layout in the extracted output to ensure that downstream processing can reference sections and tables.

You plan to call Azure Content Understanding in Foundry Tools.

You need to extract content and layout elements and detect QR codes without requiring a language model deployment.

Which built-in analyzer should you use?

1. prebuilt-documentFieldSchema
2. prebuilt-read
3. prebuilt-documentSearch
4. prebuilt-layout

Question 24

You have a Microsoft Foundry project that generates short promotional product videos.

After several clips are approved, reviewers notice a small watermark in the top-right corner of some videos.

You need to remove the watermark without regenerating the videos.

What should you do?

1. Modify the original prompt to exclude watermarks.
2. Crop the video by using the size parameter.
3. Increase the guidance scale.
4. Apply a mask-based inpainting edit to the affected part of the video.

Question 25

You have a Microsoft Foundry project that contains an agent named Agent1.

Agent runs succeed, but Foundry Control Plane does NOT display values for error rates, runs, and token usage, and the Traces tab is empty.

You need to ensure that Foundry Control Plane displays the appropriate values for Agent1.

What should you do?

1. Update Agent1 to a new version.
2. Restart Agent1 from Foundry Control Plane.
3. Assign a Log Analytics workspace to Agent1.
4. Enable Application Insights for Agent1.

Question 26

You are creating an image-editing workflow in a Microsoft Foundry project.

The workflow must meet the following requirements:

Ensure that background objects can be removed by applying a mask-based inpainting edit.

Preserve the original lighting and style of the edited images.

Use the built-in image editing controls, NOT a custom model.

You need to ensure that image edits apply exclusively inside the masked area.

How should you configure the workflow?

1. Set generation mode to image_variation and provide the original image as a reference.
2. Enable text_to_image mode and a prompt describing the desired background removal.
3. Enable image_to_image mode and a high-strength value to regenerate the full image based on the prompt.
4. Enable mask_inpainting and supply both the input image and a mask indicating which part of the image to modify.

Question 27

You have a Microsoft Foundry project that contains a customer support agent built on a deployed chat model.

The agent responses are validated by using an automated testing system that compares generated answers to stored expected outputs. Identical prompts must return consistent responses to prevent automated test failures.

You need to reduce response variability, without modifying the prompt or reducing factual accuracy.

What should you do for the model?

1. Increase the max_tokens parameter.
2. Remove stop sequences from the requests.
3. Decrease the temperature parameter.
4. Increase the temperature parameter.

Question 28

You have a Microsoft Foundry project that uses Azure AI Search to ground an agent in internal documentation.

After a recent content update, users report that the agent’s answers have become less accurate.

You need to identify whether the retrieved content is negatively influencing the model’s generated responses.

Which observability signal should you review?

1. indexer status and failure history
2. latency breakdown traces
3. prediction drift metrics
4. groundedness evaluation metrics

Question 29

Note: This section contains one or more sets of questions with the same scenario and problem. Each question presents a unique solution to the problem. You must determine whether the solution meets the stated goals. More than one solution in the set might solve the problem. It is also possible that none of the solutions in the set solve the problem.

After you answer a question in this section, you will NOT be able to return. As a result, these questions do not appear on the Review Screen.

You have a Microsoft Foundry project that contains an agent. The agent generates summaries from retrieved policy documents.

Users report that some responses omit required regulatory clauses, even when the clauses are present in the retrieved content.

You need to improve response completeness.

Solution: You increase the value of the temperature parameter.

Does this meet the goal?

1. Yes
2. No

Question 30

You plan to configure an evaluation in Microsoft Foundry for a Retrieval Augmented Generation (RAG) chat app.

You need to provide scores for groundedness, relevance, and harmful content categories.

Which two evaluation categories can you use? Each correct answer presents a complete solution.

NOTE: Each correct selection is worth one point.

1. risk and safety metrics
2. fluency evaluator
3. similarity evaluators
4. AI quality (NLP) metrics
5. AI quality (AI assisted) metrics

Question 31

Note: This section contains one or more sets of questions with the same scenario and problem. Each question presents a unique solution to the problem. You must determine whether the solution meets the stated goals. More than one solution in the set might solve the problem. It is also possible that none of the solutions in the set solve the problem.

After you answer a question in this section, you will NOT be able to return. As a result, these questions do not appear on the Review Screen.

You have a multimodal AI generative model that accepts image uploads and uses extracted image text to generate responses.

You discover that users can upload unsafe images and embed hidden instructions into images to manipulate the model.

You need to implement controls to mitigate the risk.

Solution: You configure image moderation to block unsafe content before processing the images.

Does this meet the goal?

1. Yes
2. No

Question 32

Note: This section contains one or more sets of questions with the same scenario and problem. Each question presents a unique solution to the problem. You must determine whether the solution meets the stated goals. More than one solution in the set might solve the problem. It is also possible that none of the solutions in the set solve the problem.

After you answer a question in this section, you will NOT be able to return. As a result, these questions do not appear on the Review Screen.

You have a Microsoft Foundry project that contains an agent. The agent generates summaries from retrieved policy documents.

Users report that some responses omit required regulatory clauses, even when the clauses are present in the retrieved content.

You need to improve response completeness.

Solution: You add a reflection pass that regenerates the response if the required clauses are missing.

Does this meet the goal?

1. Yes
2. No

Question 33

You have an invoice-processing application named App1 that uses Azure Content Understanding in Foundry Tools.

You are building a new Content Understanding pipeline named Pipeline1 that must meet the following requirements:

• Compare an invoice to its related purchase order

• Validate the invoice against static vendor contract documents

• Return a single structured output that includes discrepancy findings

You need to configure Pipeline1 and expose the pipeline as a single analyzer endpoint.

What should you configure?

1. a single-file task in standard mode that uses the vendor contract provided as an additional document during analysis.
2. a single-file task in standard mode that uses confidence scores enabled for the extracted fields.
3. a multiple-file task in pro mode that uses the vendor contract files as reference data
4. a multi-file task in standard mode that uses the invoice and purchase order as input to the analyzer

Question 34

You have an Azure AI Search indexer that ingests PDF policy manuals.

Client applications must display page-level citations that have bounding polygons for both text and images.

You need to add a single built-in multimodal content extraction skill to the Azure AI Search skillset. The solution must meet the following requirements:

• Provide text and image location metadata.

• Extract tables that span multiple pages.

What should you add?

1. Document Extraction
2. Azure Content Understanding in Foundry Tools
3. GenAI Prompt
4. Document Layout

Question 35

You have a Microsoft Foundry agent that grounds responses from an Azure Search index that contains the following:

• Searchable text fields for product names and product codes

• A vector field that stores embeddings for product descriptions

You need to ensure that users can query the index by using the following:

• Exact product names or codes

• Natural language descriptions of the products

What should you configure?

1. vector search only
2. hybrid search
3. keyword search only
4. semantic search only

Question 36

You have a Microsoft Foundry project that ingests scanned PDF invoices stored in Azure Blob Storage. Each invoice contains printed line items and has a table-based layout.

Extracted results are stored as structured JSON and used as grounding data for an agent in a Retrieval Augmented Generation (RAG) solution.

You need to create a single analyzer that meets the following requirements:

• Extracts the invoice number, invoice date, vendor name, and total amount across varying templates

• Returns confidence scores so that results with confidence below 0.80 can be routed for supervisor review

What should you use?

1. a Foundry agent that has groundedness guardrails enabled to extract invoice fields and confidence scores
2. a custom Azure Content Understanding in Foundry Tools analyzer that defines the required fields as the extracted fields and the returned confidence scores for routing
3. the Azure Content Understanding in Foundry Tools prebuilt-layout analyzer
4. the Azure Content Understanding in Foundry Tools prebuilt-documentSearch analyzer and search.score from the Azure AI Search results for routing

Question 37

You have a customer support agent that uses the Microsoft Foundry Agent Service.

Sometimes, customers return to a session days later to continue the same support case, and the agent must resume with the full historical context. The agent must provide the following:

• Multi-turn continuity within the session

• Cross-session continuity for the same case

• Access to the full interaction history, including user messages, agent messages, tool calls, and tool outputs

You need to ensure that the agent automatically reloads the complete history on each new turn.

What should you do?

1. Create and reuse a conversation by storing the conversation’s ID and supplying the ID on subsequent requests.
2. Persist only the final model response stored in the client application and prepend the response to future prompts.
3. Enable memory summarization on the agent definition to persist the context automatically.

Question 38

You have a Microsoft Foundry project that contains a Retrieval Augmented Generation (RAG) chat solution used by customer support agents.

You are adding an automated pre-production evaluation step to a CI/CD pipeline named Pipeline1. The evaluation will run against a labeled test dataset that contains support questions and the expected grounding context.

You need to ensure that Pipeline1 fails if unsupported content or a retrieval mismatch exceeds a defined threshold:

• responses include claims not supported by the retrieved source content

• retrieved source content does not align with the labeled expected context

Which two built-in evaluators should you use in Pipeline1? Each correct answer presents part of the solution.

NOTE: Each correct selection is worth one point.

1. Retrieval
2. Fluency
3. Coherence
4. Groundedness
5. Response Completeness

Question 39

You have a Microsoft Foundry project that contains a prompt agent used by a customer support web app.

The agent is invoked from a Python service that does NOT run in the Foundry portal.

You need to implement end-to-end tracing to capture latency breakdowns and exceptions across agent runs.

Which two components can you use? Each correct answer presents a complete solution.

NOTE: Each correct selection is worth one point.

1. a Log Analytics workspace
2. Application Insights
3. OpenTelemetry
4. the Azure Monitor Agent
5. Microsoft Sentinel

Question 40

Note: This section contains one or more sets of questions with the same scenario and problem. Each question presents a unique solution to the problem. You must determine whether the solution meets the stated goals. More than one solution in the set might solve the problem. It is also possible that none of the solutions in the set solve the problem.

After you answer a question in this section, you will NOT be able to return. As a result, these questions do not appear on the Review Screen.

You have a Microsoft Foundry project that contains an agent. The agent generates summaries from retrieved policy documents.

Users report that some responses omit required regulatory clauses, even when the clauses are present in the retrieved content.

You need to improve response completeness.

Solution: You increase the value of the max_tokens parameter.

Does this meet the goal?

1. Yes
2. No

Question 41

You have a Microsoft Foundry project that contains an agent and an image generation model deployment.

The agent generates original images from user-supplied product photos.

You need to ensure that the generated images maintain the product identity and visual characteristics of the provided photo.

What should you do?

1. Set the input_fidelity parameter to high.
2. Apply a groundedness detection filter.
3. Include a prompt and input image in the request.
4. Decrease the value of the temperature parameter.

Question 42

You have a customer support agent built by using the Microsoft Foundry Agent Service. The agent calls an Azure OpenAI model deployment.

During load testing, calls intermittently fail and return an HTTP 429 rate limit exceeded error.

You need to handle throttling to reduce call failures and improve reliability under load. The solution must remain within the service and model limits.

What should you do?

1. Create a new thread and retry the calls immediately.
2. Reduce the number of registered tools.
3. Implement a retry policy that uses exponential backoff and jitter.
4. Split uploaded content into smaller files.

Question 43

You are creating an agent workflow in a Microsoft Foundry project to support natural voice interactions.

The agent must receive continuous audio input, convert the input into text for reasoning, and then return spoken responses to a user. The workflow must meet the following requirements:

Support turn-taking dynamics, where the agent begins to generate the speech output before the user finishes speaking.

Operate with low latency to maintain conversational experience.

You need to enable both speech to text and text to speech in a real-time agent interaction.

What should you do?

1. Use batch transcription to convert the audio input and return text responses from the agent.
2. Use real-time speech to text for incoming audio and text to speech for agent responses.
3. Use an embeddings model to encode the audio, and then decode the audio into text and speech.
4. Use speech translation to convert the audio into another language and return the translated text.

Question 44

You have a Microsoft Foundry project that contains a support-ticket triage agent built by using the Foundry Agent Service.

The agent uses tools to classify the ticket type and set the ticket priority.

Sometimes, the same support case continues across multiple sessions over several days.

You need to persist state by using a durable ID to ensure that the agent can automatically reuse the full interaction history. The solution must preserve previous user messages, tool calls and tool outputs across turns and sessions.

Which runtime component should you use?

1. output item
2. agent
3. conversation
4. response

Question 45

You have a Microsoft Foundry project that contains an agent.

The agent uses Azure Content Understanding in Foundry Tools to process vendor onboarding packets. The packets include digital PDFs that contain tables and hyperlinks.

The extracted content is indexed for search and provided to a downstream agent in the Markdown format.

You need to generate a Markdown output that has a layout and a semantic structure optimized for Retrieval Augmented Generation (RAG) workflows.

Which built-in analyzer should you use?

1. prebuilt-documentFieldSchema
2. prebuilt-documentSearch
3. prebuilt-read
4. prebuilt-layout

Question 46

You have an app named App1 that uses a Microsoft Foundry multimodal model deployment.

App1 runs optical character recognition (OCR) on uploaded images and appends the OCR output to the prompt as additional context.

Some uploaded images contain embedded text.

You need to prevent potentially malicious instructions from being processed by the model.

What should you use?

1. image moderation
2. prompt shields for documents
3. protected material text
4. prompt shields for user prompts

Question 47

Note: This section contains one or more sets of questions with the same scenario and problem. Each question presents a unique solution to the problem. You must determine whether the solution meets the stated goals. More than one solution in the set might solve the problem. It is also possible that none of the solutions in the set solve the problem.

After you answer a question in this section, you will NOT be able to return. As a result, these questions do not appear on the Review Screen.

You have a multimodal AI generative model that accepts image uploads and uses extracted image text to generate responses.

You discover that users can upload unsafe images and embed hidden instructions into images to manipulate the model.

You need to implement controls to mitigate the risk.

Solution: You configure a prompt shield for documents.

Does this meet the goal?

1. Yes
2. No