Create a technical design

A logistics company uses a Dataverse-based model-driven app to manage thousands of daily shipment records. The finance department needs a nightly export of aggregated shipment data into an existing Azure SQL Database for their reporting warehouse. The export must run unattended, transform the data, and handle large batches reliably without requiring users to be signed in. As the Power Platform developer, which integration approach should you recommend?

You are designing a custom connector that a Power Automate cloud flow will use to call a third-party payroll REST API. The API requires each request to include an OAuth 2.0 access token issued by the vendor's Microsoft Entra ID tenant, and the vendor wants the connector to prompt each user to sign in with their own vendor account so that per-user audit logs are maintained. Which authentication type should you configure on the custom connector?

A backend Azure Function runs on a nightly schedule with no user present. It must write records to Dataverse using the Web API. The security team requires that the integration authenticate without storing or using any interactive user credentials, and each API call must be attributed to a non-interactive identity. What authentication approach should you design?

A model-driven app form has a lookup to the Account table. When a user selects an account, the sales rep needs the account's phone number and preferred contact method copied into fields on the current record immediately, before the record is saved. The copied values do not need to be validated or persisted server-side until save. Company policy states you should minimize both server load and licensing costs by avoiding server-side processing for purely presentational, pre-save behavior. What is the most appropriate way to implement this?

A retail company has five separate canvas apps, each maintained by a different team. All five apps need to display a standardized product-search bar with identical layout, styling, and behavior. Leadership wants a single definition of this search bar that, when updated, can be propagated to every app with minimal manual effort, while still allowing each app to pass in its own data source. Which approach best meets these requirements?

A field service team uses tablets to inspect equipment at remote sites with intermittent connectivity. Inspectors must capture data, photos, and a signature, and the app must reliably work offline and sync automatically when connectivity returns. The data is stored in Dataverse and technicians are already licensed for the Field Service and model-driven mobile experiences. As the developer, which approach best satisfies the offline-first requirement with the least custom code?

A developer must expose a complex business operation—'Recalculate Contract Pricing'—that is triggered from a model-driven form command button, from several Power Automate cloud flows, and from an external system via the Dataverse Web API. The operation must run server-side, participate in the same database transaction as other Dataverse writes, accept typed input parameters, and return a typed response. It should be callable as a single named message across all these consumers. Which Dataverse code component should the developer build?

You are designing a Dataverse solution for a conference-management app. During each event, the app writes millions of short-lived attendee session-scan records within a few hours, with unpredictable write bursts. The records are queried mainly by a single session key, require no complex relational joins or roll-up columns, and are purged shortly after each event. You must minimize write throttling during spikes and control storage cost. Which table type should you use?

A Power Platform developer must implement logic that validates and, if needed, rejects an incoming Dataverse record write before the transaction is committed to the database. The validation depends on complex server-side calculations across multiple related tables, and the user must receive an immediate error message that cancels the save if validation fails. Which implementation approach should the developer choose?

A developer builds a Power Automate cloud flow that uses the HTTP action to call an on-premises payroll REST API, plus a custom connector to a partner service. After the flow is deployed to the production environment, it fails with an error indicating that connectors cannot be used together, and the admin confirms a Data Loss Prevention (DLP) policy is applied to that environment. The HTTP action and the custom connector must both continue to work in the same flow. What should the developer recommend to the administrator?

A developer builds a Power Automate cloud flow that reads customer records from Dataverse and posts a summary message to a public Twitter account. The tenant has a data loss prevention (DLP) policy where Dataverse is classified as Business and Twitter is classified as Non-Business. When the flow is saved, it fails with a policy violation. What is the reason the flow cannot run?

A logistics company stores millions of shipment records in an external Azure SQL Database that is continuously updated by a warehouse management system. The Power Platform team must let model-driven app users view and filter these shipment records alongside native Dataverse tables, and reference them in advanced find and views. Data volume makes duplication impractical, the records must always reflect the latest source values, and users only need read access from within the app. Which approach best meets these requirements?

You are designing a Power Platform solution for a manufacturing company. Sensor devices on the factory floor stream approximately 5 million telemetry readings per day into Dataverse. The data is written far more often than it is read, individual records have a variable set of attributes depending on the sensor model, and the business needs the ingestion to scale automatically without degrading performance for the model-driven app that occasionally queries recent readings. Which type of Dataverse table should you use to store the telemetry readings?

A company uses a model-driven app built on Dataverse for case management. Sales representatives are organized into regional business units. The compliance team requires that a sales rep can only read and edit case records owned by users within their own business unit, but must never see cases from other regions. Managers, however, need read access to cases across all regions. The developer must configure this with minimal custom code. Which approach should the developer take?

Your organization has five different canvas apps that all include a customer intake form. Each form requires the same complex phone-number and postal-code validation logic, plus a consistent styled layout with an input box, label, and inline error message. Business stakeholders want the validation logic and appearance maintained in ONE place so a change propagates to every app. Which solution component should you build?

A logistics company has complex freight-cost calculation logic that must be invoked in three ways: synchronously from a model-driven form when a user clicks a button, from a Power Automate cloud flow that runs nightly, and from a plug-in that fires when an order is created. The business requires that all three callers execute the exact same server-side logic and that the logic is transactional and reusable without duplicating code. What is the best design approach?

A developer at a logistics company must build a specialized editable grid that displays shipment line items with custom color-coding, inline calculations, and a drag-to-reorder feature. This grid must be reusable across multiple model-driven app forms and must bind to a Dataverse column so administrators can add it through the form designer without writing code. Which solution component should the developer build?

A logistics company uses model-driven apps on Dataverse. A group of 12 dispatchers needs collaborative ownership of a set of 'Route' records so that any dispatcher can edit records the others create, and so that reassigning work between dispatchers requires no manual record re-sharing. Managers outside the group should have read-only access. The developer wants the lowest-maintenance approach that survives staff joining and leaving the group. Which design should the developer implement?