AB-410 — Intelligent Applications Builder Associate

Full question text, answer options, and explanations for this practice set — a spoiler-free alternative is the interactive quiz above for scored, shuffled practice.

1. 1. A business analyst needs to build a Power Platform solution that will span multiple environments (development, test, production) and must be deployable via a repeatable, versioned process. Which Power Platform construct should they use to package all the solution components together?

   • A. A SharePoint document library to store the exported Power Apps files
   • B. A Microsoft Power Platform solution to group all components (apps, flows, tables, agents) for ALM and deployment(correct)
   • C. A Dataverse environment variable stored in the default environment
   • D. A Power BI workspace that contains the related reports

   Explanation: Power Platform solutions are containers that group related components (Power Apps, Power Automate flows, Dataverse tables, agents, environment variables, etc.) for Application Lifecycle Management (ALM). Solutions can be exported as managed or unmanaged packages and imported into different environments, enabling repeatable, versioned deployments across dev/test/prod. SharePoint document libraries and Power BI workspaces do not package Power Platform components. Dataverse environment variables are components stored inside a solution, not the solution itself.

2. 2. An app builder is designing a Power Platform solution and needs to choose the right environment type. The solution will be used by the entire organization in production and requires data loss prevention policies and managed security. Which environment type should be used?

   • A. The default environment, since it is accessible by all users automatically
   • B. A sandbox environment, which provides production-like capabilities for testing
   • C. A production environment with proper DLP policies and access controls configured(correct)
   • D. A developer environment that each user creates individually

   Explanation: Production environments are the appropriate target for organization-wide business solutions. They support full capacity, data loss prevention (DLP) policies, access controls, and managed backup. The default environment is shared by everyone in the tenant and is not recommended for production workloads because it cannot be restricted adequately. Sandbox environments support testing and can be reset but are not intended for production use. Developer environments are personal, single-user environments for learning and development.

3. 3. A maker is building a Dataverse data model for a field service solution. They need to track which technician performed each service visit, and each service visit must be associated with exactly one technician. Which type of table relationship should be configured?

   • A. Many-to-many relationship between Service Visit and Technician tables
   • B. One-to-many relationship from Technician (one) to Service Visit (many), creating a lookup column on Service Visit(correct)
   • C. Activity relationship that auto-creates a timeline for each service visit
   • D. Connection relationship using the Connections entity

   Explanation: A one-to-many relationship from Technician to Service Visit means one technician can have many service visits, but each service visit is associated with exactly one technician. This relationship creates a lookup column on the Service Visit table pointing to the Technician record, which is the standard Dataverse pattern for this scenario. A many-to-many relationship would allow a service visit to have multiple technicians, which contradicts the requirement. Activity and connection relationships are specialized patterns for timeline/activity tracking and flexible associations, not direct ownership.

4. 4. A maker is configuring a Dataverse table to automatically generate a concise AI summary of a customer record's key fields whenever the record is viewed. Which column type provides this capability?

   • A. A calculated column that concatenates field values into a text string
   • B. A rollup column that aggregates related record counts
   • C. A row summary column configured with a prompt that generates an AI-powered summary of the record(correct)
   • D. A formula column using Power Fx expressions

   Explanation: Dataverse row summary columns use AI (powered by Copilot/AI Builder prompts) to generate natural-language summaries of record data, displayed in forms and views. They are distinct from calculated or rollup columns, which perform deterministic operations on field values. Formula columns use Power Fx for computed values but do not generate AI-powered summaries. Row summaries are the purpose-built feature for this scenario.

5. 5. A Power Platform architect is reviewing a proposed solution design. The solution includes a canvas app, a model-driven app, a Copilot Studio agent, and several cloud flows, all sharing the same Dataverse tables. A junior maker suggests creating a separate security role for each app. The architect disagrees. What is the correct approach to security in this scenario?

   • A. Create separate security roles for each app because Dataverse security roles are app-specific
   • B. Define Dataverse security roles based on user responsibilities and data access needs, then assign those roles to users regardless of which app they use — the same role grants access to the underlying data across all apps(correct)
   • C. Use SharePoint permission levels to control access to the Dataverse tables
   • D. Use the System Administrator role for all users to ensure no access issues arise

   Explanation: Dataverse security roles govern access to data (tables, rows, columns) — not to specific apps. Roles are defined around job functions and data access requirements, then assigned to users or teams. The same role applies whether the user accesses data through a canvas app, model-driven app, or API. Creating per-app roles (A) duplicates effort and may create inconsistencies. SharePoint permissions (C) do not apply to Dataverse. Assigning the System Administrator role to all users (D) removes all security boundaries and violates the principle of least privilege.

6. 6. A maker is analyzing business requirements for a Power Platform solution. The requirements include: a structured data capture form, approval routing, and AI-assisted data extraction from uploaded documents. Which combination of built-in Power Platform components most directly addresses all three requirements?

   • A. Canvas App for the form, Power Automate approval flow for routing, and AI Builder document processing model for extraction(correct)
   • B. Power BI report for data capture, Logic Apps for approvals, and Azure Form Recognizer for extraction
   • C. SharePoint list for data capture, Outlook rules for approval routing, and Power BI for extraction
   • D. Canvas App for the form, Teams channel for approval notifications, and a Python script for extraction

   Explanation: Canvas Apps provide flexible, custom forms for data capture. Power Automate's built-in approval actions handle structured approval routing with notifications and tracking. AI Builder's document processing models (trained or prebuilt) extract structured data from uploaded documents — all within the Power Platform ecosystem without external services. Option B uses Azure and Logic Apps, which are outside the Power Platform low-code boundary. Options C and D use non-integrated tools that don't form a cohesive Power Platform solution.

7. 7. A Dataverse table stores customer records. A maker needs to add a column that uses AI to automatically classify the sentiment of a customer's feedback text (Positive, Neutral, Negative) whenever the record is saved. Which column type enables this?

   • A. A choice column with Positive, Neutral, Negative options and a business rule to set the value
   • B. A prompt column configured with an AI prompt that analyzes the feedback text and returns a classification(correct)
   • C. A formula column using the `If()` Power Fx function to detect keywords
   • D. A rollup column that counts the number of positive words in related records

   Explanation: Prompt columns in Dataverse are AI-powered columns that execute a configured prompt (using generative AI models) against one or more fields on the record to generate a result — such as sentiment classification of feedback text. A choice column with a business rule (A) would require hard-coded keyword logic and cannot generalize like an AI model. A Power Fx `If()` formula (C) can match exact strings but not semantic sentiment. Rollup columns (D) aggregate numeric values from related records, not text sentiment.

8. 8. A maker is building a model-driven app and needs to create a new view that shows all active customer records sorted by the most recent creation date, displaying only the Name, Email, and Status columns. Where in the model-driven app designer should they configure this?

   • A. In the form designer, by adding a subgrid component
   • B. In the view designer, by creating a new view on the Customer table with the required columns and sort order(correct)
   • C. In the site map editor, by adding a new area
   • D. In the dashboard designer, by adding a list tile

   Explanation: Views in model-driven apps define how a list of records is displayed — which columns appear, how rows are filtered, and how they are sorted. The view designer allows creating new views on a table with specific column selection, filters, and sort orders. Forms define how an individual record is presented (fields layout). The site map controls app navigation structure. Dashboards display aggregated data but rely on views as data sources.

9. 9. A maker wants to build a model-driven app page that is initially designed through a natural language description rather than manual component assembly. Which feature of model-driven apps supports this AI-assisted page creation?

   • A. The classic form editor with drag-and-drop controls
   • B. Generative pages, which use natural language to create a page layout that the maker can then refine(correct)
   • C. Power BI embedded reports added as dashboard tiles
   • D. Canvas app components embedded in a model-driven form

   Explanation: Generative pages in model-driven apps allow makers to describe a page in natural language, and Copilot generates an initial page layout with components based on the description. The maker can then review and refine the result. This is distinct from the classic drag-and-drop form editor (A), which requires manual assembly. Power BI embedded tiles (C) display analytics, not app pages. Canvas components embedded in forms (D) extend model-driven apps with canvas capabilities but require manual canvas app design.

10. 10. A maker is building a canvas app that connects to a Dataverse table. They need a gallery control that automatically refreshes when a new record is added without requiring the user to manually reload the app. Which Power Fx function should be called after inserting a new record?

    • A. `Reset(Gallery1)` to clear the gallery's selection
    • B. `Refresh(Customers)` to re-fetch the Dataverse table data and update the gallery(correct)
    • C. `ClearCollect(LocalData, Customers)` called only on app start
    • D. `Navigate(Screen1, None)` to reload the screen

    Explanation: `Refresh(DataSourceName)` signals Power Apps to re-query the data source (Dataverse table) and update any controls bound to it, including the gallery. This is the correct pattern after a write operation to ensure the UI reflects the latest server-side data. `Reset(Gallery1)` clears the gallery's selected item but does not refresh data. `ClearCollect` on app start loads data into a local collection but doesn't auto-refresh after inserts. `Navigate` transitions to a screen but does not trigger a data refresh.

11. 11. A maker is building a canvas app for field technicians on mobile devices. The app must remain usable when the device is offline and sync changes when connectivity is restored. Which approach correctly implements offline capability in Power Apps?

    • A. Connect the app directly to Dataverse and rely on the Dataverse connector's built-in offline caching
    • B. Use `LoadData` and `SaveData` functions to persist data to the device's local storage, and use `Connection.Connected` to control when to sync with Dataverse(correct)
    • C. Enable the 'Offline mode' toggle in the app's Settings page, which automatically handles sync
    • D. Use a scheduled Power Automate flow to push data to the device every 5 minutes

    Explanation: Power Apps canvas apps support offline scenarios using `SaveData` (saves a collection to the device's local storage) and `LoadData` (loads it back on startup). The `Connection.Connected` signal allows the app to detect connectivity and conditionally sync local changes back to the server. The Dataverse connector does not have built-in transparent offline caching in canvas apps (A). There is no 'Offline mode' toggle in app settings (C). A scheduled cloud flow (D) requires connectivity and runs server-side, not on the device.

12. 12. A maker wants to add a conversational AI assistant to an existing canvas app so users can ask natural language questions about their data and perform guided actions. Which Power Platform capability allows adding this agent to the canvas app?

    • A. Embed a Power BI Q&A visual as an iframe inside the canvas app
    • B. Create a Copilot Studio agent and add it to the canvas app using the Copilot control or by creating a Copilot Studio agent from the canvas app(correct)
    • C. Add a chatbot component from the Power Apps component framework (PCF)
    • D. Configure a Power Automate flow that uses the OpenAI connector and triggers from a button in the canvas app

    Explanation: Copilot Studio allows building conversational agents that can be integrated into canvas apps, providing a natural language interface for querying data and performing guided actions. The `Create a Copilot Studio agent from a canvas app` feature (and the Copilot control) integrates agents natively. Power BI Q&A (A) provides natural language queries for Power BI datasets embedded in dashboards, not conversational task agents in canvas apps. There is no built-in chatbot PCF component (C). A flow with an OpenAI connector (D) can perform AI calls but does not provide a full conversational agent experience in the app.

13. 13. A maker is building a canvas app that multiple screens use to calculate a discounted price based on customer tier and product category. Instead of repeating the formula on every screen, the maker wants to define it once and reference it by name throughout the app. Which canvas app feature enables this?

    • A. A global variable set with `Set()` on app start
    • B. A named formula defined in the App's `Formulas` property using Power Fx(correct)
    • C. A collection pre-loaded with discount lookup values
    • D. A component library component with a property exposing the calculation

    Explanation: Named formulas (defined in the App object's `Formulas` property) allow makers to assign a Power Fx expression to a name, which can then be referenced anywhere in the app like a function — without storing state. Unlike global variables (A), named formulas are not stateful; they re-evaluate automatically when their dependencies change, preventing stale values. Collections (C) store tabular data, not formula logic. Component library components (D) can expose properties but are heavier constructs for UI reuse, not pure formula reuse.

14. 14. A maker has built a canvas app and needs to test it to ensure correct behavior before deploying to users. Which Power Apps tools can help diagnose and debug issues during testing? Choose 2.

    • A. The Power Apps Monitor tool, which captures network requests, formula evaluations, and control property values in real time during app sessions(correct)
    • B. The canvas app's built-in Test Studio for authoring and running automated UI test cases(correct)
    • C. GitHub Actions integration that runs canvas app tests in a CI/CD pipeline
    • D. The Dataverse audit log, which shows all canvas app formula errors
    • E. A Power Automate flow triggered on app errors that sends an email alert

    Explanation: The Power Apps Monitor (A) is a diagnostic tool that provides real-time visibility into network calls, connector operations, formula evaluation traces, and control property values — essential for debugging live sessions. Test Studio (B) is the built-in tool for writing and running automated test cases for canvas apps, enabling regression testing. GitHub Actions integration for canvas app tests (C) exists via the Power Platform CLI but is not a built-in Power Apps testing tool. The Dataverse audit log (D) records data changes, not canvas app formula errors. A flow for error emails (E) could be a monitoring pattern but is not a testing or debugging tool.

15. 15. A maker needs a Power Automate cloud flow to start whenever a new row is added to a specific Dataverse table. Which trigger type should be used?

    • A. Scheduled trigger with a recurrence interval of 1 minute
    • B. 'When a row is added, modified, or deleted' Dataverse trigger, scoped to 'Added' on the target table(correct)
    • C. HTTP request trigger that the canvas app calls manually
    • D. Manual trigger invoked from the Power Automate portal

    Explanation: The Dataverse connector's 'When a row is added, modified, or deleted' trigger fires in real time when a row event occurs on the specified table. Configuring the 'Change type' to 'Added' scopes it to new rows only. A scheduled trigger (A) runs at intervals and would require polling, introducing latency and unnecessary executions. An HTTP trigger (C) is appropriate for manually called flows, not table-driven automation. A manual trigger (D) requires a human to click run.

16. 16. A business process requires that a manager approves an employee expense report before it is posted to the accounting system. A maker is building a Power Automate cloud flow to handle this. Which action type in Power Automate is purpose-built for structured human approval tasks?

    • A. Send an email action using the Outlook connector, and wait for a reply containing 'Approve' or 'Reject'
    • B. The Approvals connector's 'Start and wait for an approval' action, which sends a structured approval request and waits for the approver's decision(correct)
    • C. A 'Delay until' action that pauses the flow for 48 hours then checks a SharePoint column
    • D. A 'Do until' loop that polls an Outlook inbox every 10 minutes for a response

    Explanation: Power Automate's Approvals connector provides structured approval actions that send adaptive card requests to approvers (via email and Teams), track responses, and return the decision and comments to the flow. This is the native, purpose-built solution for approval workflows. Parsing reply emails (A) is fragile and doesn't provide structured tracking. Polling SharePoint (C, D) are manual workarounds that are unreliable and resource-intensive. The Approvals connector handles all of this natively.

17. 17. A cloud flow processes a list of 500 customer records returned from a Dataverse query. For each customer, the flow calls an external API, updates a Dataverse row, and sends an email. The maker notices the flow is failing with a timeout error after processing about 60 records. What is the most likely cause and recommended fix?

    • A. The Dataverse query returns too many records; reduce the query to 50 records using a filter
    • B. The 'Apply to each' loop processes records sequentially by default; enabling concurrency (parallel branches) in the loop settings and reviewing connector throttling limits will improve throughput and prevent timeout(correct)
    • C. Cloud flows cannot process more than 100 actions total; split the flow into multiple smaller flows
    • D. Disable the 'Asynchronous pattern' setting on the HTTP action to prevent it from timing out

    Explanation: By default, 'Apply to each' processes items sequentially. With 500 records each requiring three actions (API call, Dataverse update, email), sequential processing can exceed the flow's run timeout. Enabling concurrency in the loop (up to 50 parallel branches) dramatically increases throughput. Additionally, connector throttling (Dataverse and email connectors have limits per minute) must be considered and may require retry policies or delays. There is no hard 100-action limit on cloud flows (A is wrong; pagination is the real data volume concern). Disabling async patterns (D) changes how HTTP responses are handled, not loop performance.

18. 18. A maker needs to build an AI Hub prompt that takes a customer complaint text as input and returns a structured JSON object with the sentiment, priority, and suggested resolution category. Which AI Hub feature allows defining and testing this prompt with input parameters?

    • A. A Power Automate flow with a 'Compose' action using a hard-coded template
    • B. An AI Builder prompt built from blank in AI Hub, with an input parameter for the complaint text, a structured output definition, and model configuration(correct)
    • C. A Dataverse formula column using the `JSON()` Power Fx function
    • D. A Power BI calculated measure that classifies sentiment using DAX

    Explanation: AI Hub in Power Platform allows makers to build custom prompts from blank or from templates, define input parameters, configure the generative AI model settings (temperature, model selection), and specify structured output formats including JSON schema. The prompt can then be consumed in canvas apps and cloud flows. A hard-coded 'Compose' action (A) does not invoke AI or accept dynamic inputs generically. A formula column (C) is for Dataverse computed fields, not AI prompt authoring. DAX (D) is an analytical query language in Power BI, not an AI prompt builder.

19. 19. A maker has built an AI Hub prompt that classifies support tickets. They need to call this prompt from a Power Automate cloud flow so that each new support ticket Dataverse row automatically gets classified when it is created. How should the prompt be consumed in the cloud flow?

    • A. Export the prompt as a ZIP file and upload it to SharePoint for the flow to read
    • B. Use the AI Builder 'Create text with GPT' action in the cloud flow and reference the saved prompt by name, passing the ticket description as the input parameter(correct)
    • C. Embed the full prompt text as a string in a 'Compose' action within the cloud flow
    • D. Use the HTTP connector to call the OpenAI REST API directly from the flow

    Explanation: AI Builder prompts saved in AI Hub are consumable in Power Automate cloud flows via the AI Builder connector's actions (such as 'Create text with GPT' or 'Run a prompt'). The maker selects the saved prompt by name and maps the flow's dynamic content (e.g., ticket description) to the prompt's input parameters. This is the native, governed, low-code pattern. Exporting to SharePoint (A) is not how AI prompts are distributed. Hard-coding the prompt text in a Compose action (C) loses the central management and versioning benefits of AI Hub. Calling the OpenAI API directly (D) bypasses AI Hub governance and requires API key management.

20. 20. A Dataverse table has a 'Contract Value' column. A maker needs to automatically calculate and store a 'Customer Lifetime Value' that sums the Contract Value of all related Contract records whenever a Contract record is saved. Which column type should be used on the Customer table?

    • A. A formula column using `Sum()` Power Fx on the related table
    • B. A rollup column configured to sum the 'Contract Value' field across all related Contract records(correct)
    • C. A calculated column referencing the Contract table with a lookup formula
    • D. A cloud flow that recalculates and updates the field on a schedule

    Explanation: Rollup columns in Dataverse are designed exactly for this scenario — they aggregate values (sum, count, min, max, average) from related records across a relationship, with optional filters. When a related record changes, the rollup column recalculates asynchronously. Calculated columns perform synchronous calculations on fields within the same record, not across related records. Formula columns (Power Fx) are for within-row computations. A scheduled cloud flow (D) introduces latency and requires maintenance.

21. 21. A model-driven app has a sales order form. The business rule is: if the order total exceeds $10,000, the 'Discount Approval Required' field must be set to 'Yes' and the 'Notes' field becomes required. A maker wants to implement this without code. Which feature is most appropriate?

    • A. A Power Automate flow triggered on row modification that sets these fields via a Dataverse update action
    • B. A Dataverse business rule configured on the Sales Order table to show/hide, set field values, and toggle required state based on the order total(correct)
    • C. A JavaScript web resource added to the form's OnLoad event
    • D. A model-driven app chart that highlights orders above $10,000

    Explanation: Dataverse business rules are configured on tables and execute in real time on forms (and optionally server-side) to set field values, show/hide fields, toggle required state, and display messages — based on conditions — without any code. They are the purpose-built, no-code solution for form logic like this. A Power Automate flow (A) runs asynchronously after saving and cannot make fields required in the UI before save. JavaScript web resources (C) require code and are harder to maintain. Charts (D) are visualizations, not business logic.

22. 22. A maker is building a cloud flow that calls a premium connector. During testing, the flow occasionally fails with a 429 (Too Many Requests) error from the connector. What is the recommended way to handle this in Power Automate?

    • A. Add a 'Delay' action of 60 seconds before every connector call to prevent throttling
    • B. Configure retry policy on the connector action (exponential backoff or fixed interval) and add error handling using 'Configure run after' with 'has failed' branches(correct)
    • C. Switch to a scheduled flow that runs only once per day to avoid rate limits
    • D. Use a 'Scope' action around the connector call and set it to ignore all errors

    Explanation: The correct pattern for handling transient throttling errors (429) is to configure the connector action's retry policy (Settings > Retry policy) to use exponential backoff, which automatically retries the call with increasing delays. For cases where retries are exhausted, 'Configure run after' allows branching on failure to implement fallback logic (e.g., logging, alerting, queueing for later). A fixed 60-second delay before every call (A) wastes time and may still not prevent throttling. Running only once a day (C) severely limits the flow's utility. A Scope set to ignore all errors (D) hides failures rather than handling them.

23. 23. A maker is connecting a Power Automate cloud flow to an external REST API that is not available as a certified connector in Power Platform. Which options allow the flow to call this external REST API? Choose 2.

    • A. Use the HTTP action (with a premium license) to make direct REST API calls with configurable method, headers, and body(correct)
    • B. Build a custom connector that wraps the external API's OpenAPI definition, then use it in the flow like any standard connector(correct)
    • C. Use the SharePoint connector's 'Send HTTP request to SharePoint' action to proxy calls to any external REST API
    • D. Embed the API's JavaScript SDK in a Power Fx formula column and call it from the flow
    • E. Export the flow to Azure Logic Apps, which supports the external API natively

    Explanation: The HTTP action (A, premium) allows flows to call any REST API by configuring the method, URL, headers, authentication, and body. Custom connectors (B) wrap an external API's definition (OpenAPI/Swagger or manual configuration) and make it available as a reusable connector with a better UX and potentially non-premium licensing for internal APIs. The SharePoint HTTP action (C) is specifically for SharePoint REST API calls, not arbitrary external APIs. Power Fx formula columns (D) are server-side Dataverse features and cannot call external REST APIs directly. Exporting to Logic Apps (E) is an extreme migration, not a solution within Power Automate.

24. 24. A global manufacturer uses a Power Platform solution to manage production work orders across 12 countries. Each country has different approval thresholds and regulatory requirements encoded in configuration. A maker is asked to ensure that updating these thresholds does not require exporting and importing a new solution version. Which Power Platform feature is designed for storing environment-specific configuration that can change without redeploying the solution?

    • A. Hard-code the values in Power Fx formulas within the canvas app
    • B. Store the values in a Dataverse configuration table that administrators update at runtime
    • C. Use Power Platform environment variables, which store configuration values that can be updated per environment without modifying or redeploying solution components(correct)
    • D. Store the values in a SharePoint list and have the canvas app read from SharePoint at startup

    Explanation: Power Platform environment variables are solution-aware configuration values that can differ between environments (dev/test/prod/per-country) and can be updated by administrators without modifying the solution components themselves. This is exactly the intended use case — separating configuration from code/logic. Hard-coding (A) requires code changes and redeployment. A Dataverse configuration table (B) works but requires building additional access controls and update UX. SharePoint lists (D) add an external dependency and require connectivity to SharePoint.

25. 25. A maker has built an AI Builder prompt that summarizes customer feedback and added it as a prompt column on the Customer Feedback Dataverse table. They now need to surface this AI summary in both a canvas app and a cloud flow. Which statements correctly describe how to consume the prompt in each context? Choose 2.

    • A. In a canvas app, the prompt column value can be read from the Dataverse table like any other column using the Dataverse connector — the AI summary is computed at the Dataverse layer and available as a field value(correct)
    • B. In a cloud flow, use the AI Builder connector's 'Run a prompt' action and reference the saved prompt by name, passing the feedback text as the input, then use the output in subsequent actions(correct)
    • C. In a canvas app, the maker must write a custom PCF control to call the AI Builder SDK and render the result
    • D. In a cloud flow, the prompt must be converted to an Azure OpenAI deployment before it can be consumed
    • E. Prompt columns are only available in model-driven apps and cannot be read from canvas apps or flows

    Explanation: Prompt columns in Dataverse compute AI-generated values at the data layer and store or return them as column values on records — making them accessible to any app or flow that reads the table (A). Canvas apps consume them through the standard Dataverse connector. In cloud flows, AI Builder prompts can be consumed via the AI Builder connector's 'Run a prompt' action (B), which calls the prompt on demand. No PCF control is required (C) — prompt column values are standard field values. Azure OpenAI deployment conversion (D) is not required; AI Hub prompts run on the platform's managed models. Prompt columns are not restricted to model-driven apps (E).