Get started with prebuilt prompts
The Power Platform empowers makers with prebuilt prompts for common business scenarios such as summarizing, classifying, extracting entities, translating, assessing sentiment, or drafting a reply. These prompts allow you to generate AI-driven responses without the need to manually create a prompt. Instead of specifying both the prompt and the text to process, you can simply provide the text you want to analyze, and the AI generates the desired output.
You can use prebuilt prompts in multiple experiences, such as:
- Dataverse low-code plug-ins
- Power Apps formulas
- Power Automate
Prerequisites
An environment in a region where prompts are available.
To learn if prompts are available in your region, go to Feature availability by region—Prompts.
A Power Apps or Dynamics 365 license.
A Dataverse database installed on the environment.
An AI Builder add-on.
Available prebuilt prompts
The following tables describe the available prebuilt prompts. If you find a prebuilt prompt you want to use, go to Use a prebuilt prompt in this article to learn how to use it. If you don't find a prebuilt prompt that meets your needs, you can create a custom prompt.
AISummarize
Summarize the text that you provide. For example, this prompt summarizes an email message or text from a document.
Input for AISummarize
| Name | Required | Type | Description | Values |
|---|---|---|---|---|
| Text | Yes | string | Text to summarize | Text sentences |
Output for AISummarize
| Name | Description | Values |
|---|---|---|
| SummarizedText | Summarized version of the input text | Text sentences |
AISentiment
Detect the sentiment of the text that you provide. For example, this prompt detects whether the sentiment of a customer review is positive, negative, or neutral.
Input for AISentiment
| Name | Required | Type | Description | Values |
|---|---|---|---|---|
| Text | Yes | string | Text to analyze | Text sentences |
Output for AISentiment
| Name | Description | Values |
|---|---|---|
| AnalyzedSentiment | Overall sentiment of the analyzed text | Positive, neutral, or negative |
AIReply
Draft a reply to the message that you provide. For example, this prompt drafts a reply to a customer's review of a product.
Input for AIReply
| Name | Required | Type | Description | Values |
|---|---|---|---|---|
| Text | Yes | string | Text to respond | Text sentences |
Output for AIReply
| Name | Description | Values |
|---|---|---|
| PreparedResponse | A draft message in response to the provided input text | Text sentences |
AITranslate
Translate text from another language. For example, this prompt translates a customer email or product review. The source language doesn't need to be specified, and is automatically detected.
To learn more about languages supported for the source and target language, go to Translator language support—Translation and review the list of languages in the Auto Language Detection column.
Input for AITranslate
| Name | Required | Type | Description | Values |
|---|---|---|---|---|
| Text | Yes | string | Text to translate | Text sentences |
| TargetLanguage | No | string | Language code to which you want to translate | A valid language code such as en for English |
Output for AITranslate
| Name | Description | Values |
|---|---|---|
| TranslatedText | Translated text | Text sentences |
AIClassify
Classify text into one or more provided categories. For example, the following list of categories might be used to classify issues submitted by your customers:
["Problem", "Billing", "How To", "Licensing"]
Input for AIClassify
| Name | Required | Type | Description | Values |
|---|---|---|---|---|
| Text | Yes | string | Text to classify | Text sentences |
| Categories | Yes | table (string array) | Categories | Table of categories |
Output for AIClassify
| Name | Description | Values |
|---|---|---|
| Classification | Category | Name of selected category |
AIExtract
Extract specified entities such as registration numbers, phone numbers, or names of people.
Input for AIExtract
| Name | Required | Type | Description | Values |
|---|---|---|---|---|
| Text | Yes | string | Text from which to extract data | Text sentences |
| Entity | Yes | string | Entity to extract | Name of entity to extract |
Output for AIExtract
| Name | Type | Description | Values |
|---|---|---|---|
| ExtractedData | table | Extracted data that matched the type of entity provided | Table of zero or more rows of data matching the provided entity. |
Use a prebuilt prompt
The following sections provide links for how to use a prebuilt prompt in a Dataverse low-code plug-in, a Power App, or a Power Automate flow.
Use in a Dataverse low-code plug-in
To learn how to use prebuilt prompts in a low-code plug-in, go to Use a prebuilt prompt in a Dataverse low-code plug-in.
Use in Power Apps
Except for AI Translate, each of the prebuilt prompts mentioned previously can be used in Power Apps the same way you can use custom prompts. To learn how to use prompts in Power Apps, go to Use your prompt in Power Apps.
Use in Power Automate
Except for AI Translate, each of the prebuilt prompts mentioned previously can be used in Power Automate the same way you can use custom prompts. To learn how to use prompts in Power Automate, go to Use your prompt in Power Automate flow.
Provide feedback
If you have feedback of your experience with the current prebuilt prompts, or ideas for more prebuilt prompts, you can share your thoughts on the Power Apps Ideas portaL. To do so, select the AI Builder checkbox in the Categories list in the Filters sidebar.
See also
Feedback
Coming soon: Throughout 2024 we will be phasing out GitHub Issues as the feedback mechanism for content and replacing it with a new feedback system. For more information see: https://aka.ms/ContentUserFeedback.
Submit and view feedback for