# AB-100 Agentic AI Business Solutions Architect practice exams

> Prepare for AB-100 with 240 practice questions across 5 full timed exams, weighted toward Plan AI-powered business solutions. Every answer is explained.

Page: https://devopscodelab.com/exams/microsoft-ab-100/

240 questions in 5 timed practice tests. Each test has about 50 questions and a 100-minute limit. The pass mark is 70%. Results are split across the 3 exam domains.

## Exam at a glance

| Setting | Value |
| --- | --- |
| Practice tests | 5 |
| Questions in total | 240 |
| Questions per test | About 50 |
| Time limit per test | 100 minutes |
| Pass mark | 70% |
| Question types | Single answer and multiple answer (24 multiple-answer questions) |
| Level | Professional |
| Category | AI & Machine Learning |

These are the settings of our practice tests. Check the official Microsoft exam page for the current format before you book.

## Domains and weights

| Domain | Weight | Questions in our bank |
| --- | --- | --- |
| Plan AI-powered business solutions | 41% | 98 |
| Design AI-powered business solutions | 35% | 84 |
| Deploy AI-powered business solutions | 24% | 58 |

## Price and access

- AB-100 practice exams: $14.99 for 12 months of access.
- Free mini exam: 15 questions with explanations (free account): https://devopscodelab.com/app/exams/microsoft-ab-100/mini/
- All-Access subscription includes every exam: https://devopscodelab.com/pricing/
- 10 free questions with answers: https://devopscodelab.com/exams/microsoft-ab-100/free-practice-questions/

## Sample questions

### Question 1 of 3

A bank wants customer service, fraud and lending agents to collaborate. The customer-facing agent must run in Microsoft Teams and on the website, while the fraud agent needs custom Python code, a fine-tuned model and private networking to internal scoring APIs. Which design BEST fits these requirements?

- A. Build the customer-facing agent in Copilot Studio and connect it to a fraud agent hosted in Microsoft Foundry Agent Service as a connected agent
- B. Build all three agents as declarative agents in Microsoft 365 Copilot and share a single SharePoint knowledge source across all of them
- C. Build the customer-facing agent as a Foundry agent and expose it to website visitors through a Microsoft 365 Copilot declarative agent
- D. Build all three agents in Copilot Studio as child agents of one parent and call the scoring APIs through public HTTP request nodes

**Answer:** A

**Explanation:** Copilot Studio provides low-code multichannel publishing to Teams and websites and can call Foundry agents, which support custom code, fine-tuned models and private networking for the fraud workload. Declarative agents only run inside Microsoft 365 Copilot and cannot host custom Python or reach private scoring APIs. Child agents in a single Copilot Studio agent do not give the fraud team code-first control or private networking, and declarative agents are not a channel for anonymous website visitors.

### Question 2 of 3

A retailer plans to ground a Copilot Studio agent on its product catalog. The catalog lives in three systems: a Dataverse table updated hourly, a SharePoint library of PDF spec sheets last reviewed in 2021, and an ERP export that contains duplicate SKUs. Which data-readiness issue should the architect flag as the HIGHEST risk to answer quality?

- A. The Dataverse table is refreshed hourly, so the agent may return stock levels that are up to sixty minutes behind the warehouse system
- B. Using three different repositories means the agent must use classic orchestration because generative orchestration supports one source
- C. The SharePoint spec sheets have not been reviewed for years, so the agent may confidently ground answers on outdated product details
- D. The ERP export is a flat file, so the agent cannot read it until every record has been migrated into a SharePoint list with metadata

**Answer:** C

**Explanation:** Timeliness is a core grounding-data criterion, and documents that have not been reviewed in years are the most likely source of confidently wrong answers. An hourly Dataverse refresh is acceptable for most catalog questions. Flat files can be cleaned and loaded into a supported source, and generative orchestration can use multiple knowledge sources, so neither of those is the main risk; the duplicate SKUs should also be cleaned but are secondary to stale content.

### Question 3 of 3

An HR team wants employees to ask questions about policies stored in SharePoint from inside Microsoft 365 Copilot Chat. They need no custom code, no external channels and must respect existing SharePoint permissions. What should the architect recommend?

- A. A Power Apps canvas app with an embedded AI Builder prompt that reads the SharePoint list through a premium HTTP connector
- B. A custom engine agent built with the Microsoft 365 Agents SDK, hosted in Azure App Service and grounded through Azure AI Search
- C. A declarative agent created with Copilot Studio or Agent Builder that uses the HR SharePoint site as its knowledge source
- D. A Foundry agent that uses a fine-tuned GPT model trained on exported policy documents and published through a Teams app manifest

**Answer:** C

**Explanation:** Declarative agents extend Microsoft 365 Copilot with instructions and knowledge, inherit the user's SharePoint permissions and require no code, which matches every stated requirement. A custom engine agent or a Foundry agent adds hosting, code and indexing work that the scenario does not need. A canvas app with a prompt is a separate app experience rather than an extension of Copilot Chat and would need extra security design.


## Study plans

### Two weeks

1. Day 1: take practice test 1 in exam mode to get a baseline.
2. Days 2 to 5: domain mode on Plan AI-powered business solutions and Design AI-powered business solutions, the two most heavily weighted domains.
3. Day 6: practice test 2 in exam mode, then the mistakes quiz.
4. Days 7 to 12: practice tests 3 to 5, one every other day, each followed by its mistakes quiz.
5. Days 13 and 14: final mode. Book the exam when your readiness stays above the pass mark.

### Four weeks

1. Week 1: one practice test in practice mode, reading every explanation. Note your two lowest domains.
2. Week 2: domain mode on those two domains, then a second practice test in exam mode.
3. Week 3: practice tests 3 and 4 in exam mode, with the mistakes and flagged quizzes after each.
4. Week 4: practice test 5 and final mode until your readiness holds above the pass mark.

## Questions about the AB-100 practice exams

**Are these real AB-100 exam questions?**

No. We write every question ourselves and map it to the published AB-100 exam objectives. We do not use leaked or recalled questions.

**Do the time limit and pass mark match the official exam?**

They match the settings of our practice tests. Exam providers change their format from time to time, so check the official exam page before you book.

**Can I try it before I pay?**

Yes. The free mini exam has 15 questions with full explanations and needs no account.

**How long do I keep access?**

A single-exam purchase gives you 12 months of access. All-Access includes this exam for as long as your subscription runs.

**What happens after I finish a test?**

You get a score, a pass or fail against the pass mark, a score per domain and every answer with its explanation.

Last updated November 2026.

DevOpsCode Lab is not affiliated with Microsoft. AB-100 and related names are trademarks of their owners. We write every question ourselves; these are not real exam questions.
