top of page

An Overview of AI Builder for Power Platform

Updated: Mar 30, 2023

Image courtesy of Microsoft

To help you enhance your business processes, AI Builder delivers a new feature for Microsoft Power Platform that automate processes and anticipate outcomes based on input data. There are four primary products in Microsoft's Power Platform: Power BI, Power Apps, Power Automate, and Power Virtual Agents.

We can use Power Automate and Power Apps to bring intelligence to our company using AI Builder, and we don't need to know how to code to do it. A wide range of AI models are available from various AI builders, and we may apply them in Power Automate. You may use AI Builder to build your own models or train an existing model to meet your company's specific requirements.

Using workflow automation, Power Automate from Microsoft can turn any repetitive and time-consuming operation into an automated process. Workflows may be created by SharePoint experts to send emails, gather data, alerts, extract data from multiple inputs such as emails and documents, etc.

It's possible to analyze and save data in a database or SharePoint list, for example, using AI Builder. It's possible to use categories like "negative," "positive," or "excellent" to sort through and classify customer comments and reviews.

AI Model Types

AI builder allows you to pick pre-built AI models that can be used in Power Automate or Power Apps without having to train and construct the model as appropriate for your business needs. Custom AI Models and prebuilt AI Models make up the bulk of it. Custom models may also be developed and trained.

Custom AI Models

1. Prediction: Using yes/no patterns derived from past/historical data, the prediction model enables us to forecast future events in yes/no patterns. For example, who could be able to take the test?

2. Form processing: You may use machine learning to evaluate and extract key-value pairs from document and picture files. A minimum of five papers is required to begin. By using training data from documents, you may teach the computer model to pull out specific information from those papers.

3. Object detection: You can speed up or automate business operations by using object detection. Managing inventory and customer interactions, for example, may be made easier with its assistance.

4. Category classification: The Classification model categorizes massive amounts of text data, such as emails, social media posts, papers, etc., so that it may be extracted and acted upon as appropriate. This model can recognize text items that have a variety of tags attached to them. Spam detection, for example, or redirecting customer requests.

5. Entity extraction: The entity extraction model enables you to identify the data according to your business requirements. Unstructured data is transformed into machine-readable structured data by identifying and classifying important parts. An example of this is "SharePoint Architect," which will be labelled "Designation" in your data.

Pre-built AI Models

1. Business Card Reader: Business Card Reader lets you extract data from photos of business cards. For example, a Name, Job Title, Address, and Email may be extracted from the picture.

2. Text Recognition: The Text Recognition model enables you to extract words from photos and texts into a machine-readable format. Extract any keyword from a document, for example.

3. Category Classification: The Category Classification model enables you to categorize full text into categories with distinct tags. Some pre-built models for client feedback, such as Issues and Compliance as well as Customer Service and Documentation as well as Pricing and Billing, are available.

4. Key Phrase Extraction: Key Phrase Extraction enables you to extract keyphrases (main points) from the text. The words "restaurant" and "great staff" will be extracted from a sentence such, "The restaurant was nice. There was amazing staff."

5. Language Detection: It is possible to determine the language of a written document using Language Detection. A statement like "I have an automobile" will return "en" rather than "en-US." as an example. Unknown will be the result if the language detection fails.

Sentiment Analysis: It's possible to identify positive, negative, natural, and mixed feelings in text using the Sentiment Analysis model. It is possible to get a label of Positive when you say, "I am pleased while working at GemRain".

AI Builder in Power Automate
AI Builder in Power Automate

AI Builder usage in Power Apps
AI Builder usage in Power Apps


These are courses that our Master Trainer developed throughout his years of experience in training the official Microsoft courseware. He had compiled all questions and feedbacks commonly asked in all of his different pieces of training and crafted this 100% hands-on style workshop series. This ensures all attendees share the same real-world experiences that our Master Trainer evangelized all these years. Most of our clients thoroughly enjoyed this variety of training and had broadly endorsed them. Our Custom-Developed Course (CDC) courses are conducted in a workshop format, which is training based entirely on instructor-led hands-on driven classes. Our trainers thoroughly guide and perform hands-on lab exercises with the attendees to give them the maximum learning experience possible.

If you have any questions regarding the training, please don't hesitate to contact us directly at


bottom of page