AB-100 Practice Questions
Prepare for AB-100 with more than an answer.
- Exam fee
- $165 USD
- Level
- Advanced (Expert badge)
- Valid for
- 12 months (renewal frequency)
Domains covered on the exam 3
- Plan AI-powered business solutions27.5%
- Design AI-powered business solutions27.5%
- Deploy AI-powered business solutions42.5%
- 1
Your company uses a legacy mainframe terminal application that lacks any APIs, database access, or modern integration points. You need a Copilot Studio agent to check inventory status by navigating this terminal application.
Which TWO features or approaches must you design into the solution to achieve this? (Select TWO)
Show answer details
Correct answer: A, E
When an application lacks APIs or standard connectors, you must use UI-based automation. In the context of advanced agents, 'Computer Use' allows the agent to visually interpret a screen, click, and type to automate tasks. This is typically implemented alongside Power Automate Desktop (RPA) which acts as the execution engine for these UI interactions.
When an application lacks APIs or standard connectors, you must use UI-based automation. In the context of advanced agents, 'Computer Use' allows the agent to visually interpret a screen, click, and type to automate tasks. This is typically implemented alongside Power Automate Desktop (RPA) which acts as the execution engine for these UI interactions.
- 2
True or False: According to the Microsoft Power Platform Well-Architected Framework for intelligent application workloads, you should always bypass user-level permissions and grant AI components elevated system access to ensure maximum data retrieval efficiency.
Show answer details
Correct answer: B
False. The Well-Architected Framework emphasizes security as a foundational pillar. AI components and agents must respect user-level permissions and role-based access control (RBAC). Granting elevated system access violates the principle of least privilege and introduces massive security and data leakage risks.
- 3
You are designing a Copilot Studio agent for an HR department. If a user asks a question that the agent cannot answer using its configured knowledge bases or custom topics, the agent currently responds with a generic error message and drops the conversation.
Which component must you configure to handle these unrecognized intents gracefully and escalate to a human HR representative?
Show answer details
Correct answer: B
In Copilot Studio, the Fallback topic is a system topic that is triggered when the agent's natural language understanding cannot match the user's input to any existing topic or knowledge base. Configuring the Fallback topic allows you to provide a graceful response, ask clarifying questions, or initiate an escalation to a human agent.
- 4
You are orchestrating AI features in Dynamics 365 Finance. Users want to be able to ask the in-app Copilot questions about specific corporate financial policies that are not part of the standard Microsoft documentation.
Which action should you recommend to enable this capability?
Show answer details
Correct answer: C
Dynamics 365 Finance and Supply Chain Management allow administrators to extend the in-app Copilot help experience by adding custom knowledge sources (like SharePoint sites containing corporate policies). This grounds the finance and operations agent chats on your specific organizational data.
- 5
You are designing a Power Apps canvas app for field technicians. Instead of building complex UI screens manually, you want the app interface to dynamically generate UI elements based on the data structure and user context using AI.
To achieve this, you should propose the use of code-first ___________ pages and an agent feed.
Show answer details
Correct answer: C
Code-first generative pages (and the use of an agent feed) allow developers to embed AI components directly into Power Apps canvas apps. This enables the UI to adapt dynamically and generate responses or interfaces based on AI orchestration, rather than relying solely on static, pre-built screens.
- 6
Alpine Ski House is planning a comprehensive AI adoption strategy using the Cloud Adoption Framework (CAF) for AI. The organization has identified several potential use cases for Copilot Studio agents and wants to prioritize them based on technical feasibility and business value. During which phase of the CAF AI adoption process should the team build a proof of value (PoV) for the highest-priority agent?
Show answer details
Correct answer: B
In the Cloud Adoption Framework for AI, the 'Plan' phase involves evaluating technical feasibility, selecting the first workloads, and building a proof of value (PoV) or proof of concept. The Strategy phase is for defining business justification and expected outcomes. The Ready phase focuses on preparing the environment (landing zones), and the Adopt phase is where full-scale development and deployment occur.
- 7
Woodgrove Bank is designing an AI-powered document processing solution. They need to analyze thousands of documents daily. Some requests require complex reasoning and synthesis, while others only require basic summarization and entity extraction. The Chief Architect wants to optimize the total cost of ownership (TCO) without sacrificing accuracy on complex tasks.
Which TWO architectural components should you recommend implementing to achieve this cost optimization? (Select TWO)
Show answer details
Correct answer: A, E
To optimize TCO, a model router dynamically directs incoming prompts to the most appropriate model based on complexity, intent, or cost parameters. By pairing this with a mix of Large Language Models (LLMs) for complex tasks and Small Language Models (SLMs) for basic tasks, the organization avoids paying premium LLM costs for simple summarization.
To optimize TCO, a model router dynamically directs incoming prompts to the most appropriate model based on complexity, intent, or cost parameters. By pairing this with a mix of Large Language Models (LLMs) for complex tasks and Small Language Models (SLMs) for basic tasks, the organization avoids paying premium LLM costs for simple summarization.
flowchart LR Req[User Request] --> Router{Model Router} Router -->|Complex Reasoning| LLM[GPT-4o] Router -->|Basic Summarization| SLM[Phi-3 Mini] LLM --> Resp[Response] SLM --> Resp - 8
You are assessing Dataverse tables that will be used to ground a new Dynamics 365 Customer Service agent. During your review, you discover that the 'KnowledgeArticle' table contains thousands of duplicate entries, HTML formatting errors, and conflicting resolution steps for the same error codes.
Which data quality dimension for grounding is primarily failing in this scenario?
Show answer details
Correct answer: C
Cleanliness refers to the absence of duplicates, formatting errors, and conflicting information in the dataset. Poor cleanliness directly impacts an agent's ability to provide coherent and accurate responses when grounding on that data. Timeliness refers to data being up-to-date, relevance refers to the data matching the domain, and availability refers to the system's ability to access the data.
- 9
CASE STUDY
Company Background:
Fabrikam is a global manufacturing enterprise that heavily utilizes Dynamics 365 Sales and Microsoft 365. They have a proprietary, highly complex pricing engine hosted on-premises that calculates custom discounts based on real-time supply chain metrics, historical customer purchases, and current commodity prices.Current Situation:
Sales representatives currently switch between Microsoft Teams, Dynamics 365 Sales, and a legacy desktop application to calculate prices during customer calls. This context switching increases call times and leads to calculation errors.Requirements:
- Sales reps must be able to ask natural language questions in Microsoft Teams (e.g., "What is the maximum discount I can offer Northwind for 500 units of Product X today?").
- The solution must securely query the on-premises pricing engine.
- The solution must understand complex manufacturing terminology and Fabrikam's unique discount tiering logic.
- The solution should leverage existing Microsoft 365 Copilot investments where possible to minimize licensing costs.
flowchart TD User[Sales Rep in Teams] --> Copilot[Microsoft 365 Copilot] Copilot --> ?[Missing Component] ?[Missing Component] --> OnPrem[(On-Premises Pricing Engine)]Based on Fabrikam's requirements, which architectural approach represents the optimal build vs. buy vs. extend decision?
Show answer details
Correct answer: B
Extending Microsoft 365 Copilot using a declarative agent with an API plugin is the optimal choice. It meets the requirement to surface in Teams, leverages existing M365 Copilot investments (avoiding the cost of building a fully custom UI/orchestrator), and can securely access the on-premises pricing engine via a standard API plugin routed through an on-premises data gateway. Building a fully custom agent ignores the requirement to leverage existing M365 investments.
- 10
Lucerne Publishing is establishing a Microsoft AI Center of Excellence (CoE) to govern their rollout of agentic solutions across Dynamics 365 and Power Platform. The Chief Information Officer wants to ensure the CoE focuses on the correct foundational pillars.
Which of the following is a primary element of the Microsoft AI Center of Excellence strategy that the team must implement?
Show answer details
Correct answer: C
A Microsoft AI Center of Excellence (CoE) focuses on establishing governance, sharing best practices, providing reusable assets (like templates and prompt libraries), and ensuring responsible AI adherence across the organization. It is meant to empower fusion teams, not bottleneck all AI development to a single centralized developer team.
