Get Your CertAI-103PlayCompete
‹ AI-103 tests

Microsoft Azure AI Apps and Agents Developer Associate (AI-103) practice questions

100 questions in 10 tests. Open one to answer it on its own, or take a whole test.

Test 1

Take Test 1 ›
  1. A support organization runs one job every night that sends roughly two million closed tickets through a chat model to classify them by root cause. Nobody is waiting on the output… · Plan and manage an Azure AI solution
  2. Users of a customer-facing chat feature report that it occasionally returns an error during short traffic spikes, then works normally again a minute later. Application logs show… · Plan and manage an Azure AI solution
  3. A pharmaceutical company runs an internal assistant on a Foundry model deployment. Legal has supplied a list of about 60 unreleased drug codenames and an internal merger project… · Plan and manage an Azure AI solution
  4. An HR team wants an internal assistant that answers benefits questions. The requirements are modest: a system prompt, one deployed model, the file search tool over the employee… · Implement generative AI and agentic solutions
  5. A prompt agent in production has four tools attached, and one of those tools calls a second internal service of its own. Over the past week its 95th-percentile response time has… · Implement generative AI and agentic solutions
  6. A logistics company runs a prompt agent in Foundry Agent Service that customers reach through a chat widget. The agent must look up live shipment status from an existing internal… · Implement generative AI and agentic solutions
  7. A furniture retailer has 4,000 existing studio photographs of sofas. Marketing wants to swap the plain grey backdrop in each photo for a styled living-room scene while leaving… · Implement computer vision solutions
  8. A broadcaster is indexing a 20,000-hour advertising archive so that an internal agent can answer questions like "show me the clips where our client's logo appears next to a… · Implement computer vision solutions
  9. A contact centre stores the previous day's chat and voice transcripts in an overnight job so analysts can study them. Compliance requires that customer names, phone numbers, and… · Implement text analysis solutions
  10. A RAG assistant grounded on 60,000 chunks of internal policy prose is retrieving the wrong passages in two distinct ways. Queries that quote an exact clause identifier such as… · Implement information extraction solutions

Test 2

Take Test 2 ›
  1. An order-management microservice needs to call a published prompt agent's endpoint in a Microsoft Foundry project. Security review requires that the service principal be able to… · Plan and manage an Azure AI solution
  2. A Python web app running on Azure App Service currently calls chat completions with an API key read from Key Vault. The team is moving the app onto Foundry Agent Service, and… · Plan and manage an Azure AI solution
  3. A procurement assistant answers questions over supplier-submitted PDFs that are ingested nightly into a search index. During a penetration test, a consultant uploaded a contract… · Plan and manage an Azure AI solution
  4. A claims service prompts a chat model to return a claimant name, a policy number, and an array of line items, then calls json.loads on the content and indexes the fields… · Implement generative AI and agentic solutions
  5. A platform team supports fourteen agents: some are Foundry prompt agents, the rest are code agents built on Agent Framework and LangGraph. All fourteen wire up the same six… · Implement generative AI and agentic solutions
  6. Before launch, reviewers of a RAG assistant flag a recurring problem: the answers read fluently, address the question that was asked, and cite the right policy area, but they… · Implement generative AI and agentic solutions
  7. A bank is building a multi-turn assistant on Foundry Agent Service. Its regulator requires that no conversation content be persisted by the service: every message must live only… · Implement generative AI and agentic solutions
  8. A news agency is retrofitting alt text onto a photo archive. For each photograph the accessibility team wants a single sentence describing the picture as a whole, plus separate… · Implement computer vision solutions
  9. A newsroom uploads finished 40-minute interview recordings and an editor waits at the keyboard for the transcript before cutting subtitles. Nothing is live; each file is complete… · Implement text analysis solutions
  10. An ingestion pipeline is working: an Azure AI Search indexer runs a skillset with the Text Split skill and the AzureOpenAIEmbedding skill, and the vector field in the index is… · Implement information extraction solutions

Test 3

Take Test 3 ›
  1. A support agent in Foundry Agent Service reads ticket text through an MCP tool that pulls from a third-party helpdesk. The team assigned a guardrail to the model deployment with… · Plan and manage an Azure AI solution
  2. A government contractor has deployed a Foundry resource with public network access disabled, a private endpoint for inbound traffic, and virtual network injection for the agent's… · Plan and manage an Azure AI solution
  3. A support assistant handles a lopsided mix of traffic: most turns are one-line FAQ lookups, and perhaps one in six is a genuinely hard multi-step troubleshooting question. Every… · Plan and manage an Azure AI solution
  4. A platform agent is being connected to an internal MCP server that exposes forty tools. Six are read-only lookups the agent genuinely needs; the rest include operations such as… · Implement generative AI and agentic solutions
  5. An HR prompt agent answers policy questions from the employee handbook, which is uploaded to the project and served through the file search tool. A new requirement lands: the… · Implement generative AI and agentic solutions
  6. A legal team's drafting assistant has two distinct problems. First, it must quote clause language from a precedent library that is amended most weeks. Second, every output must… · Implement generative AI and agentic solutions
  7. A marketing team wants short product teaser clips generated from a written brief plus a reference product photograph. After review, the usual note comes back: one element of an… · Implement computer vision solutions
  8. A second-hand marketplace lets sellers upload listing photographs. Before a photo goes live, the platform must screen it for sexual, violent, hateful and self-harm content. The… · Implement computer vision solutions
  9. A life-sciences company translates regulatory submissions from English into German. The general Translator model is fluent but renders the company's controlled process and… · Implement text analysis solutions
  10. A health insurer receives referral letters from several hundred clinics. They are free-text, multi-page and narrative; no two practices lay one out the same way. For each letter… · Implement information extraction solutions

Test 4

Take Test 4 ›
  1. A retailer runs its assistant on a Global Provisioned deployment sized for normal load. During flash sales the PTUs saturate and callers start receiving 429 responses. Finance… · Plan and manage an Azure AI solution
  2. A prompt agent works correctly in the agents playground, where it reads reference documents from a storage account through a configured tool. The team saves it as a version and… · Plan and manage an Azure AI solution
  3. A bank is choosing an environment configuration for Foundry Agent Service. Its data-governance policy is short and non-negotiable: every file an agent ingests, every conversation… · Plan and manage an Azure AI solution
  4. A finance team wants an agent that takes an uploaded CSV of monthly results, computes exact year-on-year variances per cost centre, and returns a bar chart image the analyst can… · Implement generative AI and agentic solutions
  5. A booking agent is being hardened before release. Trace review shows a consistent pattern: the agent reliably picks the correct tool for each request and never calls a tool it… · Implement generative AI and agentic solutions
  6. A company already runs a production Azure AI Search index over its knowledge base. Two other applications query it, an indexer keeps it current on a schedule, and the relevance… · Implement generative AI and agentic solutions
  7. A support assistant sends a 4,000-token system message (product catalogue and answering policy, unchanged for months) on every request. To help the model with relative dates, the… · Implement generative AI and agentic solutions
  8. Claims adjusters at an insurer want to ask open-ended questions about damage photographs: "is the rear bumper cracked or only scuffed?", "is there a second vehicle in the… · Implement computer vision solutions
  9. A retailer wants its voice assistant to speak with a distinctive, recognizable brand voice recorded by an actor the company has contracted, rather than with one of the standard… · Implement text analysis solutions
  10. A RAG assistant grounds on documents stored in Azure Data Lake Storage Gen2, where access is already governed by ACLs and RBAC assignments on Microsoft Entra groups. Today the… · Implement information extraction solutions

Test 5

Take Test 5 ›
  1. A consulting firm's platform team is standing up Foundry for several client-facing delivery teams. Every team already has Azure Machine Learning prompt-flow pipelines it must… · Plan and manage an Azure AI solution
  2. A retailer's platform team has already created a Foundry project and deployed the models it uses. A new developer joins to build and test prompt agents inside that project… · Plan and manage an Azure AI solution
  3. A financial services company runs two Foundry model deployments: one behind an external customer chatbot, one behind an internal research assistant. After an incident review, the… · Plan and manage an Azure AI solution
  4. A retailer's shopping-assistant prompt agent already carries a large toolbox for inventory lookup, order status, and recommendations. A separate fraud team owns and independently… · Implement generative AI and agentic solutions
  5. A support ticketing system tags every incoming ticket with one of five fixed categories (billing, technical, account, shipping, other) before routing it. The classification… · Implement generative AI and agentic solutions
  6. An application defines get_weather and get_traffic as tools on a chat completions deployment. A user asks: "What's the weather and traffic like in Seattle and in Denver?" The… · Implement generative AI and agentic solutions
  7. A retail app lets shoppers photograph a shelf price tag. The app must return the printed price and product code as text so it can compare them against the online catalog. The… · Implement computer vision solutions
  8. A neobank's mobile onboarding flow captures a live selfie to verify a new customer's identity against their uploaded ID photo. During testing, a tester defeats the check by… · Implement computer vision solutions
  9. A hotel chain analyzes guest reviews with Azure AI Language sentiment analysis. One review reads: "The room was spotless and quiet, but the front-desk staff were rude and… · Implement text analysis solutions
  10. A logistics company receives purchase orders as PDF and image attachments from about 30 different suppliers. Every PO carries the same kind of information: PO number, supplier… · Implement information extraction solutions

Test 6

Take Test 6 ›
  1. A bank disabled public network access on its Foundry resource and added a private endpoint on a subnet in its hub virtual network. The private endpoint connection shows Approved… · Plan and manage an Azure AI solution
  2. An architect is designing a customer-facing chat feature around a single Global Standard model deployment in one Foundry resource. To satisfy a business continuity requirement… · Plan and manage an Azure AI solution
  3. A developer provisions a new Standard (pay-as-you-go) deployment of a chat model. In the Foundry portal, the deployment's capacity is expressed as a tokens-per-minute (TPM)… · Plan and manage an Azure AI solution
  4. A support chat widget calls a chat completions endpoint and waits for the entire answer before showing anything. Longer answers leave the widget blank for several seconds, and… · Implement generative AI and agentic solutions
  5. A multi-turn support app calls the Chat Completions API directly. On every new user turn, the app reloads the entire prior transcript from its own database and resends it as the… · Implement generative AI and agentic solutions
  6. A team classifies incoming support tickets with a chat model. The system message only says "Classify the ticket into one of: Billing, Technical, Account." Responses are… · Implement generative AI and agentic solutions
  7. A RAG pipeline embeds entire ten-page manual sections as single chunks. Retrieval usually returns a section that does contain the relevant sentence, but each chunk is so large… · Implement generative AI and agentic solutions
  8. A stock-photo marketplace has hundreds of thousands of untagged photos. It wants shoppers to type a free-text description, such as "a red kayak on a calm lake at sunset," and get… · Implement computer vision solutions
  9. A media-monitoring service scans thousands of news articles a day and wants a short list of the main topics each article covers, for a headline-browsing screen: for example, an… · Implement text analysis solutions
  10. An accounting team receives scanned supplier statements. Every supplier lays its statement out differently, but each one contains an itemized table of charges that must load into… · Implement information extraction solutions

Test 7

Take Test 7 ›
  1. A two-person team is prototyping an internal tool that only calls Azure AI Language for key phrase extraction and Azure AI Vision for OCR on scanned forms. They want a single key… · Plan and manage an Azure AI solution
  2. A platform team runs a shared Foundry resource that several application teams call. After an incident where nobody could say which application caused a spike in 429 errors, the… · Plan and manage an Azure AI solution
  3. A hospital network's compliance team requires that it be able to revoke the encryption key for its Foundry resource's data at rest at any time, cutting off access immediately. An… · Plan and manage an Azure AI solution
  4. A creative-writing assistant lets users type free-form prompts directly to the model; it does not retrieve any documents or call any tools. During a red-team exercise, a tester… · Implement generative AI and agentic solutions
  5. A developer builds a support chatbot by calling the Azure OpenAI chat completions API directly: each turn, the app appends the new user message to a list and resends the entire… · Implement generative AI and agentic solutions
  6. A retailer already embeds and stores product-catalog vectors in its own Azure Cosmos DB account, where several other services query them with the database's native vector index… · Implement generative AI and agentic solutions
  7. A quality-control team wants a custom model that, given a photo of a pallet holding several mixed products, returns the location of each individual item on the pallet along with… · Implement computer vision solutions
  8. Eighteen months ago, a developer built a product-photo pipeline that called the Image Analysis 4.0 Segment API's background removal operation to cut product photos onto… · Implement computer vision solutions
  9. A support platform receives tickets in whatever language the customer typed, with no language field set by the submission form. Before a ticket reaches Custom Translator or a… · Implement text analysis solutions
  10. A logistics company built three Document Intelligence custom models, one each for its purchase orders, its bills of lading, and its customs declarations, each a distinct… · Implement information extraction solutions

Test 8

Take Test 8 ›
  1. A claims-processing service calls the Document Intelligence analyze REST endpoint from a .NET backend. Six months ago a developer wired the call to whatever api-version the… · Plan and manage an Azure AI solution
  2. A German insurer wants to deploy a newly released Azure OpenAI model. Company policy requires that all customer data stay within the EU. The insurer's usual region, Germany West… · Plan and manage an Azure AI solution
  3. A US hospital network's compliance policy states that patient data sent to a generative AI model must never be processed outside the United States, full stop. Their current… · Plan and manage an Azure AI solution
  4. A developer is preparing a training file to run a supervised fine-tuning job against a chat model in Microsoft Foundry. They have 200 example support exchanges and want to format… · Implement generative AI and agentic solutions
  5. A team runs a customer-facing chatbot on a Global Provisioned (PTU) deployment of a specific GPT model version. Six weeks ago they received the model's retirement-date notice by… · Implement generative AI and agentic solutions
  6. A search team is indexing 50 million document chunks in Azure AI Search and wants to shrink vector storage as much as possible using binary quantization together with Matryoshka… · Implement generative AI and agentic solutions
  7. A RAG chatbot is evaluated for groundedness and passes consistently; every answer is traceable to the retrieved context with no fabricated content. Manual review turns up a… · Implement generative AI and agentic solutions
  8. A media company lets editors upload a single photo and needs the app to automatically generate several thumbnail variants from it at different aspect ratios (a 1:1 square tile… · Implement computer vision solutions
  9. A compliance team needs to summarize lengthy regulatory filings. Their auditors require that every sentence appearing in a summary be an exact, unaltered quote lifted from the… · Implement text analysis solutions
  10. A law firm stores scanned contract PDFs, image-only pages with no digital text layer, in Blob Storage and wants an Azure AI Search index where every document has filterable… · Implement information extraction solutions

Test 9

Take Test 9 ›
  1. A team is building a prompt agent in Foundry Agent Service that must ground its answers in live public web results and return the bing_grounding tool's source citations for a… · Plan and manage an Azure AI solution
  2. A nightly batch pipeline runs forty parallel worker processes that all call the same Azure AI Foundry model deployment. Each worker already implements exponential backoff on 429… · Plan and manage an Azure AI solution
  3. A platform team is building a GitHub Actions workflow that runs on every merge to main to deploy and update a Foundry resource and its model deployments. Security requires that… · Plan and manage an Azure AI solution
  4. A developer's polling loop reads run.status in a tight loop and treats any value other than completed as "still running," so it just sleeps and checks again, including when the… · Implement generative AI and agentic solutions
  5. A regulated fintech company runs Foundry Agent Service with the standard agent setup. Compliance requires that all customer conversation history be stored exclusively in Azure… · Implement generative AI and agentic solutions
  6. A support chatbot has a get_order_status function tool defined on every request. For messages where the customer types "agent," the team wants that specific request to guarantee… · Implement generative AI and agentic solutions
  7. A user-generated video platform needs to automatically flag uploaded clips that contain adult or racy visuals, and separately flag clips whose spoken transcript contains explicit… · Implement computer vision solutions
  8. A city transit agency wants to automatically blur faces in dashcam footage before releasing it publicly, using only face detection bounding boxes; it never needs to identify who… · Implement computer vision solutions
  9. A pharmaceutical support line transcribes calls in real time. Everyday conversation transcribes accurately, but drug names and internal SKU codes are consistently misheard… · Implement text analysis solutions
  10. An enterprise runs an Azure AI Search indexer and skillset over scanned invoices, extracting key phrases, entities, and structured line-item tables. Separately, the data… · Implement information extraction solutions

Test 10

Take Test 10 ›
  1. A logistics company's dispatch assistant runs around the clock and answers drivers in real time. Request volume is unusually stable (close to the same number of requests per… · Plan and manage an Azure AI solution
  2. A platform team is standardizing Microsoft Foundry usage across dozens of subscriptions owned by different business units. Leadership's rule: only models published directly by… · Plan and manage an Azure AI solution
  3. A team is writing a disaster-recovery plan for its Foundry Agent Service deployment. They've learned that the service keeps no built-in backups and supports no point-in-time… · Plan and manage an Azure AI solution
  4. An operations team wants an agent to submit maintenance tickets in a decades-old internal web portal. The portal exposes no API of any kind, only a browser-rendered UI, so no… · Implement generative AI and agentic solutions
  5. A bank's customer-support web app lets a signed-in customer chat with an agent about their own accounts. To answer "what's my current balance," the agent needs an internal… · Implement generative AI and agentic solutions
  6. A team runs a Hosted agent in production; version 3 currently serves 100% of the endpoint's traffic. They want to test version 4, a new container image with an updated tool… · Implement generative AI and agentic solutions
  7. An applied-AI team wants to compare two prompt variants by running each against 50,000 questions from a fixed test set (100,000 requests in total) one time only, to decide which… · Implement generative AI and agentic solutions
  8. A stock-photo marketplace wants to auto-generate searchable keywords for every photo it hosts, dozens of loosely related single words per image, such as "beach," "outdoor,"… · Implement computer vision solutions
  9. A hospital's automated phone line reads medication instructions aloud using text-to-speech. Testers report that a caution ("take only one tablet, not two") is read in the same… · Implement text analysis solutions
  10. An accounts-payable team ingests two kinds of paper: point-of-sale receipts employees submit for expense reimbursement (merchant name, date, tax, tip, total), and vendor bills… · Implement information extraction solutions