How to Create an App and Data Model with Natural Language in Power Apps
A shipping company is evaluating Microsoft Power Platform services and capabilities. The company requires an app that can track and manage stock items. The company has no existing data or images to assist with app creation. You need to demonstrate the ease of the process to create an app and a data model by using natural language. Which platform feature or option should you use for the demonstration?
Community Insight
Tests Copilot's ability to generate both a Power Apps app and Dataverse tables from a conversational description; the trap is choosing 'Start with data' or 'Start from an image', which require pre-existing assets the company lacks.
In Power Apps, Copilot uses natural language to generate an app and its underlying Dataverse data model without existing data or images. This page establishes that Copilot (A) is the correct PL-900 answer for demonstrating rapid, natural-language app and data-model creation.
The most common wrong answer is 'Start with data' (B) because it sounds like a logical data-first approach, but it requires an existing data source and cannot create the app and data model from natural language.
Community Discussion (1 comments)
Comments & Corrections
No comments yet — spotted an error or have a note? Share it below.
Expert Analysis
Why the Answer Is Correct
Copilot in Power Apps is the only option that matches the scenario's demand for natural-language creation of both an app and a data model. The question explicitly states the company has no existing data or images, which rules out approaches that depend on those assets. Copilot lets a maker describe the desired stock-tracking app in plain language, and it generates a working canvas app together with an underlying Dataverse table structure. As Cooljoy7777 noted, Copilot is the natural language assistant that can "generate both a Power Apps application and its underlying Dataverse table structure." That directly satisfies the requirement to demonstrate the ease of creating an app and a data model.Why the Other Options Are Wrong
"Start with data" (B) requires you to connect to or import an existing data source, so it cannot work when the company has no data. "Start from an image" (C) needs an existing image to seed the app, which the company also lacks. Power Automate (D) is a workflow automation service, not a tool for generating an app or its data model from natural language. AI Builder (E) adds AI capabilities such as form processing or prediction to an app, but it does not create the app scaffolding or Dataverse tables from a conversational description. Each of these options fails at least one explicit constraint in the question.Community Comment Notes
The lone community comment, from Cooljoy7777, aligns with the official Power Apps Copilot documentation and emphasizes that Copilot understands conversational descriptions. It points out that Copilot generates both the application and its underlying Dataverse table structure, which is exactly the dual deliverable the question asks for. No dissenting opinions appear in the comment thread, and the comment's reasoning reinforces why "Start with data" and "Start from an image" are traps. The comment is thus a reliable confirmation that A is the intended answer.Exam Strategy
When a PL-900 question emphasizes natural language and the absence of existing data or images, look for Copilot as the answer. It is the only option that can generate both the app and its data model from a description, so eliminate options that require pre-existing assets.
Frequently Asked Questions
Why is 'Start with data' not the right choice for this stock app?
It requires an existing data source to generate the app; the scenario states the company has no existing data, so it cannot demonstrate natural-language app and data-model creation.
Can AI Builder generate the stock-tracking app from a natural language description?
No, AI Builder builds AI models such as form processing or prediction; it does not generate an app or Dataverse table structure from natural language.
Related Analysis
Practice All PL-900 Questions
Access 130 questions with complete answers and detailed explanations.
View Full PL-900 Practice Test →