Skip to main content

Power Apps is empowering coders with next-generation AI capabilities

Headshot of article author Ryan Cunningham

Every year at Microsoft Build, Microsoft returns to its roots and puts developers at center stage. This year, Microsoft Power Apps is announcing a powerful set of next-generation AI capabilities and developer-focused features that improve productivity for all developers—including those who code.

In March, we announced a next-generation AI Copilot in Power Apps that will transform low-code development, building on the ability to generate Microsoft Power Fx from natural language and create apps from uploaded images. For the first time, Copilot in Power Apps gave developers the ability to create complex data tables and single-screen applications all from natural language.

Today, we take the next step in the journey of AI-assisted development:

  • Copilot can create multi-screen applications and make real-time adjustments with the power of natural language.
  • Copilot is adding data processing capabilities by helping you clean and prepare your data before you even start building the app. Copilot can now ingest any Excel file, no matter how loosely structured, and create robust structured tables with a variety of data types.
  • Copilot control now brings the power of intelligence to your favorite model apps with its intuitive data understanding and convenient navigation. Copilot control in canvas apps also gets a major boost in understanding your Dataverse data.
  • Copilot is making life easier for developers and admins alike with the ability to generate descriptions for complex applications and multi-component solutions with apps, data tables, and automations.

While Copilot is changing everything about the way we work, it’s not the only update we’re announcing to improve the lives of developers everywhere. One of the most important tools for any developer is the ability to reuse the components they create. Experienced developers accelerate their productivity by creating robust libraries of modules they can quickly deploy in new applications. Now, Microsoft Power Platform is enabling organizations to tap into those robust libraries with organization-wide component catalogs. These catalogs allow anyone to submit and install approved components that other developers at their company have created. This catalog is key to enabling any organization to scale its low-code practice company-wide.

Copilot can now create complex multi-screen applications

While the initial iteration of Copilot in Power Apps helped you get started with an app over Microsoft Dataverse, now you can go much further working side-by-side with AI. Makers have already seen the power of Copilot in creating and editing the data tables that underpin their applications. Now, Copilot will be able to create complex multi-screen applications through natural language conversation and design and update beautiful user interfaces. All of this means less time spent on mundane tasks like aligning buttons and more time spent on solving complex coding challenges.

Here are just a few examples of natural language prompts you can use:

  • “Add a screen where users can submit and rank their favorite ideas from my existing screen templates.”
  • “Add a button to my page that allows them to sort their data.”
  • “Change the color of all my app’s buttons to blue.”
  • “Center all buttons in their containers and turn them orange.”

Anyone with a Dataverse environment in the United States can get started using Copilot in Power Apps today.

Copilot control now enables developers to add the power of Copilot to their own applications

Copilot isn’t just for makers. With copilot control, we gave developers the ability to add the power of Copilot into their custom apps. Now, copilot control will be available to all Power Apps makers, whether they are building model-driven applications or custom apps. Copilot in model-driven applications understands and can reason over the Dataverse tables in your app—large or small and can even navigate through the data in your app. All of this is wrapped up in a new copilot control user experience with a brand-new look that follows you where you need it most. This copilot can maintain multi-turn conversations to navigate and derive insights about your application’s data.

Here are just a few examples of natural language prompts you can use:

  • “How many high-priority tickets were created last month?”
    • “What is the breakdown of the current status for the tickets that were created last month?”
    • “Which types of tickets are most overdue?”
  • “What are the most common budget items where actual spend exceeds estimated spend?”
  • “During which months do our sales geographies generate the highest revenue?”
    • “How does that change for just our West Coast geographies?”
  • “Take me to my support agents page.”

Learn more about how copilot control is putting the power of generative AI in every application to drive insights for users.

Copilot can now understand unstructured file inputs and turn them into structured data tables

To help speed up and simplify the creation of tables and apps from Excel, makers can now drop their Excel files into Power Apps, see a quick preview of the table that will be created and a preview of the data that will be imported. Makers can tweak the names, descriptions, data types, and other table properties before creating a brand-new app that is automatically connected to the new table and data.

Makers with Copilot enabled will get a helping hand in the creation of a great table for their data. Copilot will review the incoming data to come up with appropriate column names, ensure the right data types are selected, and even convert simple columns into choices if it sees a collection of repeating values. Copilot will also provide names and a description for the table based on the data. This can all still be achieved with Excel files that don’t group cells into a table and can even deal with tables that have no column headers present.

Learn more about how Copilot in Power Apps is automating the process of data preparation and ingestion.

Copilot can generate descriptions of your applications and solutions in seconds

Admins today rely on app descriptions to manage the hundreds of apps within their organization. Without descriptions, they have no way to know what the apps do without opening each one individually. This makes it hard to scale. Coming soon, Copilot will be able to help you generate a description of an app by ingesting the app’s metadata and returning an overview of what the app does in natural language. This alleviates the need for developers to manually write summaries and for admins to open individual apps every time they need to understand what each app does.

Solutions are complex, often containing dozens of components that are difficult to partition and summarize. In the case of pipeline deployments, admins and approvers currently must open every component in every solution to gain an understanding of what they do. By using AI to help generate useful solution descriptions, developers save valuable time and get high-quality descriptions of all their apps and solutions.

More information about Copilot’s ability to generate descriptions of applications and solutions coming soon!

Sharing reusable development components to achieve organizational level-impact

Modern software development works by building on components, packages, and libraries that other developers created as foundational blocks. A developer’s productivity is a function of how robust and efficient their tools are. With better tools comes faster iteration. With reusable tools comes scalability. As developers and administrators look to expand the reach and impact of their low-code applications, the need for a central repository of low-code components that anyone in an organization can leverage is clear. That’s why we’re announcing a brand-new catalog in Microsoft Power Platform that gives makers a place to publish, share, and reuse the best low-code components within their organization. We are also empowering administrators to ensure only components that meet the highest standards of quality and security are published through the catalog manager.

Submitting a component to the library is simple and requires just a few lines of code in the command line interface:

Once the component is submitted, the environment administrator is prompted to review and either approve or reject the component.

Installation of components from the catalog is easy in just a few lines using the command line interface.

Once you install the component, it will appear as a control in the maker studio. You can easily drag and drop it into your application to start building on the work of your organization’s best code-first developers.

Learn more about the new catalog in Microsoft Power Platform.

These advancements—significant improvements to Copilot and the release of the catalog—will accelerate the productivity of every professional developer who is building on Power Apps. Join the era of AI-assisted development and see how generative AI can amplify your impact.

Get started with next-generation AI capabilities

  1. Learn more about the latest next-generation AI updates to Copilot.
  2. Submit your first components to the catalog in Microsoft Power Platform.
  3. Need a free developer plan? Get started at Power Apps Developer Plan.

Learn about Microsoft AI and other announcements across Microsoft Power Platform:

Availability 

These features are currently available in the United States and English language by signing up. Note that these preview features are experimental, and accuracy and relevance cannot be guaranteed.

Responsible AI 

Microsoft is committed to creating responsible AI by design. Our work is guided by a core set of principles: fairness, reliability and safety, privacy and security, inclusiveness, transparency, and accountability. We are helping our customers use our AI products responsibly, sharing our learnings, and building trust-based partnerships. For these new services, we provide our customers with information about the intended uses, capabilities, and limitations of our AI platform service, so they have the knowledge necessary to make responsible deployment choices.