AB-100
Free Microsoft Certified: Agentic AI Business Solutions Architect Practice Test
12 exam-style questions with full explanations — no sign-up. Score yourself, then close your gaps with the full course.
Exam fee ~$1653 exam domainsLevel Advanced2 timed practice tests in the course
✅ Free practice — no sign-up📝 Real exam-style questions💡 Detailed explanations💸 30-day money-back via Udemy
A solution architect is designing a multi-agent system where a coordinating agent must break a complex customer request into subtasks, delegate each subtask to specialized agents with different knowledge domains, and then synthesize their individual outputs into a single final answer, with the plan adjusting dynamically as results come in. Which orchestration pattern BEST fits this requirement?
Magentic orchestration (based on Microsoft Research's Magentic-One approach) uses a manager agent that dynamically plans, assigns subtasks to specialized agents, and aggregates/synthesizes their outputs, adjusting the plan as new information arrives - exactly the scenario described.
A retail company wants to prioritize 12 candidate agent use cases before starting a pilot. Leadership wants to focus first on projects that deliver measurable value quickly while requiring the least engineering effort. Which prioritization approach BEST fits this requirement during the envisioning phase?
A value-versus-effort (impact/feasibility) matrix is the standard envisioning technique to identify quick wins that maximize ROI while minimizing implementation risk, which is exactly what the stakeholders are asking for.
An organization has built a multi-agent solution in Copilot Studio and Microsoft Foundry that is fully tested in a development environment. They need to promote it to production with version history, rollback capability, and mandatory approval gates before each release. Which approach BEST satisfies these ALM requirements?
Power Platform pipelines (or Azure DevOps/GitHub Actions with the Power Platform Build Tools) package agent solutions as managed solutions, provide versioning, promote artifacts sequentially across dev/test/prod, and support approval gates before deployment — the recommended ALM pattern in AB-100.
An architect is designing an agentic solution and needs to explain the difference between Model Context Protocol (MCP) and the Agent2Agent (A2A) protocol to stakeholders. Which statement correctly describes their primary difference?
MCP defines a standard interface for connecting AI models/agents to external tools, data sources, and context providers. A2A defines a standard for cross-vendor, cross-platform communication and task delegation between independent agents. The two are complementary and are often used together in the same architecture.
An architect is scoping a customer-service transformation initiative. The business requirement is a single conversational agent that answers FAQs and escalates unresolved tickets to a human, with no need to call multiple specialized agents or external systems beyond one CRM connector. Which planning conclusion is correct?
When requirements describe a single agent with one connector and simple escalation, a single-agent Copilot Studio solution is architecturally appropriate; over-engineering with multi-agent orchestration adds unnecessary complexity and cost, and scope can expand later.
A Foundry-based customer service agent needs to detect and block jailbreak and prompt injection attempts embedded in user input before that input reaches the underlying model. Which Microsoft capability should be configured to address this specific risk?
Prompt Shields, part of Azure AI Content Safety, is purpose-built to analyze inbound prompts and detect both direct jailbreak attempts and indirect prompt injection hidden in documents or grounding data before the model processes them.
A Copilot Studio agent must answer questions using only content from a specific SharePoint site and must never fall back to general public/web knowledge when generating an answer. Which configuration should the architect apply?
By default, Copilot Studio generative answers can blend knowledge-source content with general/public knowledge. To strictly ground responses to only the configured SharePoint content, the general knowledge/public content option must be explicitly turned off in the generative answers configuration.
During the definition of success metrics for an agentic AI initiative, a stakeholder asks for a metric that reflects business outcome improvement AFTER the agent has been in production for several months, as opposed to a metric that predicts early adoption trends. Which term describes the outcome-based metric being requested?
Lagging indicators measure outcomes after they occur (e.g., cost savings, resolved-ticket rate over months), confirming whether the solution achieved its business goal, which is what the stakeholder is describing.
An architect must prevent a Copilot Studio agent from combining data retrieved from Dataverse with results from a public web search connector within the same conversation turn, because they belong to different data sensitivity tiers. What should be configured to enforce this separation?
DLP policies in Power Platform classify connectors into groups (Business, Non-Business, Blocked) and prevent connectors from different groups from being used together in the same app or agent flow, which is exactly the mechanism for isolating Dataverse data from a public web search connector.
A sales team wants an agent embedded in Dynamics 365 Sales that can create and update records, and the same capability must also be available in a custom Teams app built with Copilot Studio, without duplicating logic. Which approach is recommended?
Building the capability once as a reusable Copilot Studio agent backed by Dataverse/Power Platform connectors, then registering it as a connected (child) agent, allows both the Dynamics 365 copilot and the Teams app to invoke the same logic without duplication, consistent with agentic composition patterns on the Microsoft agent stack.
When defining the ROI model for a proposed HR onboarding agent, which pair of metrics should be captured in the business case to satisfy the AB-100 planning methodology for measuring success?
A defensible ROI/business case requires a documented baseline (current cost/time) compared against projected savings after deployment, with clear ownership for tracking — this is the core planning artifact tested on AB-100.
A retail company has deployed several Copilot Studio agents and wants to track containment rate, session volume, and estimated cost savings per resolved conversation to justify continued investment to leadership. Which approach BEST supports this ongoing ROI reporting need?
Copilot Studio provides native analytics (sessions, engagement, resolution/escalation rates) and can be integrated with Application Insights for deeper telemetry; exporting this data into Power BI enables custom ROI dashboards tracking containment and cost-per-resolution over time — the standard operationalization pattern tested on AB-100.
Ready for the real thing?
The full course has two full-length practice tests, video lessons for every exam domain, hands-on labs and detailed answer explanations.
Start my full course on Udemy →AB-100 exam — quick answers
How much does the AB-100 exam cost?
The exam fee is approximately $165 and varies by region — confirm current pricing with the certification vendor before you book.
What topics are on the exam?
It covers 3 domains: Design AI-powered business solutions (~34%), Plan AI-powered business solutions (~33%), Deploy AI-powered business solutions (~33%). The full course has a dedicated chapter, lab and practice-test coverage for each.
Is this practice test really free?
Yes — all questions on this page are free with explanations and no sign-up. The paid Udemy course adds two full-length timed exams, video lessons and hands-on labs.
Will this prepare me for the real exam?
The questions mirror the real exam's style and are mapped to the official domains. This is exam-focused preparation — combine the free test with the full course's timed simulations to gauge your readiness.