The Power Platform’s AI Builder empowers users to integrate artificial intelligence directly into business applications—even if you’re not a developer. In this article, we’ll explore what AI Builder is, review its out‐of‐the‐box prompts, walk through how to customize one (using an example of extracting key information from resumes), and explain how to leverage your custom prompt in both Power Automate and Power Apps.

What Is AI Builder?

AI Builder is Microsoft’s low-code solution for embedding AI capabilities into your apps and flows. It provides a library of prebuilt prompts—such as text analysis, summarization, sentiment detection, classification, reply, and extraction—that allow you to quickly address common business scenarios. For more background on AI Builder, check out Microsoft’s Learn AI Builder documentation .

Exploring Out-of-the-Box Prompts

When you open Power Apps or Power Automate, you’ll notice an “AI Hub” option that gives you immediate access to the available prompts. You can view the complete library, which is organized into categories like language and text analysis, content creation and management, and tasks and roles for specific tools. Among these, five automated prompts are ready to use:

https://learn.microsoft.com/en-us/ai-builder/media/power-automate-ai-models.png

AI Classify
AI Summarize
AI Sentiment
AI Reply
AI Extract

These prebuilt models give you a head start—but if you have a unique scenario (for example, extracting factual data from resumes), you can easily customize a prompt.

Customizing a Prompt: Extracting Information from Resumes

Imagine you’re recruiting a new team member and have dozens of resumes to review. You might need to pull out key details—certifications, past job experiences, education, and more—from each document. In our example, we start with the “Extract Information from Text” prompt and then modify it to work with files (like resumes) instead of plain text. Key changes include:

Changing the Input Source:
Instead of text, switch the prompt input to accept a file (e.g., a PDF resume).
Defining Entities:
Customize the entities (certifications, previous jobs, experience, education) so the AI knows exactly which information to extract.
Adjusting Output Format:
Decide whether you want the extracted information in plain text or JSON format. In our demo, the output is in JSON, which makes it easy to integrate with downstream processes.

For additional guidance on prompt customization—including tweaking settings such as the model used, temperature (to control creativity vs. predictability), and record retrieval options—see Microsoft’s detailed guidance on AI Builder prompt model settings .

Leveraging Custom Prompts in Power Automate

Once your custom prompt is ready, you can easily integrate it into a Power Automate flow. Here’s a high-level walkthrough:

Trigger the Flow:
Use a SharePoint list (for example, a job application tracker) as your flow’s trigger. When a new item is created (with an attached resume), the flow begins.
Retrieve the Attachment:
Use the SharePoint “Get attachments” and “Get attachment content” actions to fetch the resume file.
Call the AI Builder Prompt:
In Power Automate, use the action called “Create text with GPT using prompt.” When you add this action, your custom prompt appears in a dropdown list.
Process the Output:
The response—formatted in JSON—is then parsed using a “Parse JSON” action. You can extract individual elements (like certifications) and store or use them as needed.
Update SharePoint:
Finally, update the SharePoint list with the extracted information so you can easily compare candidates side by side.

This approach automates the extraction process and reduces manual data entry, all while harnessing the power of AI.

Leveraging Custom Prompts in Power Apps

In Power Apps, you can integrate your custom AI Builder prompt into a canvas app. The process differs slightly from Power Automate:

Connect Your Prompt:
Within the Power Apps Studio, add a data source and search for your custom prompt (e.g., “Extract Information from Resume”).
Trigger the AI Function:
On a button’s “OnSelect” property, call the prompt using the appropriate Power Fx function. You’ll pass in the file attachment (from SharePoint, for example) and the relevant entity details.
Display the Results:
Once the AI Builder extracts the data, use galleries or collections to display the information in your app—showing certifications, experience, education, etc.

This integration makes it easy for users to interact with the AI without leaving the app, providing a seamless and intuitive experience.

Important Note on Credit Consumption

It’s worth mentioning that while testing your prompt within the AI Builder interface does not consume credits, integrating it into Power Automate or Power Apps does count toward your credit usage. Always be mindful of this when designing your solutions.

Conclusion and Next Steps

By following these steps, you can quickly move from exploring built-in AI prompts to customizing a prompt that suits your business needs—like extracting key information from resumes. Whether you’re automating a recruitment process in Power Automate or building a dynamic interface in Power Apps, AI Builder is a powerful tool that makes advanced AI accessible to everyone.

For further learning and detailed tutorials on AI Builder prompts, consider checking out the Microsoft training module on Grounded Prompts .

Start exploring AI Builder today and unlock new possibilities for automating and enhancing your business processes!

AI Series: Getting Started with AI Builder in Power Platform: Build and Leverage Custom Prompts