Using AI Builder in Canvas Apps – A Comprehensive Guide
Introduction
AI Builder is a Power Platform capability that allows users to integrate AI models into their Power Apps without requiring complex coding or data science expertise. By using AI Builder in Canvas Apps, we can add AI-powered functionalities such as text recognition, object detection, sentiment analysis, business card reading, form processing, and prediction models.
This guide will provide a detailed step-by-step explanation of how to use AI Builder in Canvas Apps, covering setup, model selection, integration, testing, and optimization.
1. Understanding AI Builder in Power Apps
1.1 What is AI Builder?
AI Builder is a low-code AI solution that enables users to create and deploy AI models within Power Apps and Power Automate. It simplifies machine learning and AI integration for business applications.
Key Features of AI Builder:
✔ Prebuilt AI models for common business scenarios
✔ Custom AI models for specific business needs
✔ Seamless integration with Power Apps, Power Automate, and Dataverse
✔ No-code/low-code AI model training
✔ Supports image, text, and data processing
2. Setting Up AI Builder in Power Apps
2.1 Checking AI Builder Licensing and Availability
Before using AI Builder, ensure that your Power Apps environment has access to AI Builder credits. AI Builder is available in Power Apps Premium or as an add-on license.
🔍 To check AI Builder availability:
1️⃣ Sign in to Power Apps (https://make.powerapps.com)
2️⃣ Click on Settings (⚙️) > Plan(s) & Billing
3️⃣ Look for AI Builder Credits availability
💡 Note: If you don’t have AI Builder credits, contact your administrator or get a free trial from Microsoft.
3. Choosing and Training an AI Model in AI Builder
3.1 Accessing AI Builder Models
🔍 Steps to Open AI Builder:
1️⃣ Go to Power Apps Studio
2️⃣ Click on AI Builder (left navigation panel)
3️⃣ Select Build to see available models
3.2 Types of AI Models Available
AI Builder provides two main types of models:
A) Prebuilt AI Models (No training required)
🔹 Business Card Reader (Extracts information from business cards)
🔹 Sentiment Analysis (Analyzes positive/negative sentiment in text)
🔹 Text Recognition (Extracts text from images)
🔹 Language Detection (Identifies the language of text)
🔹 Key Phrase Extraction (Extracts important phrases from text)
B) Custom AI Models (Requires training)
🔹 Object Detection (Detects objects in images)
🔹 Form Processing (Extracts text from structured forms)
🔹 Prediction Model (Predicts outcomes based on past data)
💡 Prebuilt models can be used instantly, whereas custom models require training before use.
3.3 Training a Custom AI Model
🔍 Steps to Train an AI Model:
1️⃣ Select a model type (e.g., Object Detection)
2️⃣ Click Create and provide a name for the model
3️⃣ Upload sample data (images, forms, or datasets)
4️⃣ Label the data (e.g., tag objects in images)
5️⃣ Click Train Model
6️⃣ Once training is complete, publish the model
💡 Model training time depends on the dataset size. You can track progress in the AI Builder dashboard.
4. Adding AI Builder to a Canvas App
4.1 Creating a Canvas App
🔍 Steps to Create a Canvas App:
1️⃣ Open Power Apps Studio
2️⃣ Click Create → Canvas App
3️⃣ Choose Blank App (Phone or Tablet layout)
4️⃣ Click Create
4.2 Inserting an AI Builder Control
After creating the Canvas App, we need to add AI Builder components to the app interface.
🔍 Steps to Add an AI Builder Control:
1️⃣ Click on Insert (top menu)
2️⃣ Select AI Builder from the list
3️⃣ Choose the AI model control (e.g., Business Card Reader, Object Detector)
4️⃣ Resize and position the control in the app
4.3 Connecting an AI Model to the App
If using a custom AI model, we need to connect it to the app.
🔍 Steps to Connect AI Model:
1️⃣ Select the AI Builder component
2️⃣ Go to the Properties panel
3️⃣ Click AI Model and choose the trained model
4️⃣ Save and refresh the app
5. Using AI Builder Controls in Power Apps
5.1 Example 1: Using Business Card Reader
🔍 Steps to Implement Business Card Reader:
1️⃣ Insert Business Card Reader control
2️⃣ Add a Gallery to display extracted data
3️⃣ Set Gallery’s Items Property to:
BusinessCardReader1.Data
4️⃣ Add Text Labels to display details like:
First(Label).BusinessCardReader1.FullName
First(Label).BusinessCardReader1.Email
5️⃣ Run the app and test with a business card image
💡 The app will automatically extract and display contact details.
5.2 Example 2: Using Object Detection
🔍 Steps to Implement Object Detection:
1️⃣ Insert Object Detector control
2️⃣ Upload images with identifiable objects
3️⃣ Set DetectedObjects property to:
ObjectDetector1.DetectedObjects
4️⃣ Display object details using Gallery and Labels
💡 AI Builder will recognize and list objects in the image.
5.3 Example 3: Using Form Processing
🔍 Steps to Implement Form Processing:
1️⃣ Train a Form Processing model using sample forms
2️⃣ Insert Form Processor control in the Canvas App
3️⃣ Set Form Processor’s Document property to:
FormProcessor1.ExtractedData
4️⃣ Display extracted form fields in Labels or Galleries
💡 The app will extract text from structured forms like invoices or receipts.
6. Optimizing AI Builder Performance
6.1 Best Practices for AI Builder in Canvas Apps
✔ Use prebuilt models when possible to reduce setup time
✔ Train AI models with high-quality data for better accuracy
✔ Limit AI calls to improve app performance
✔ Use Power Automate for complex AI tasks
✔ Store AI results in Dataverse to minimize repeated calls
7. Testing and Publishing the AI-Powered App
7.1 Testing AI Features
🔍 Steps to Test AI Builder in Power Apps:
1️⃣ Click Preview App (Play button)
2️⃣ Upload images or input text to test AI functionality
3️⃣ Verify AI model outputs and accuracy
7.2 Publishing the AI-Integrated App
🔍 Steps to Publish the App:
1️⃣ Click File > Save
2️⃣ Click Publish to Organization
3️⃣ Share the app with users
💡 Ensure AI Builder is enabled for all users before publishing.
8. Conclusion
🔹 AI Builder empowers Power Apps with intelligent automation.
🔹 Users can leverage prebuilt and custom AI models for enhanced app functionality.
🔹 Canvas Apps can integrate AI to analyze text, detect objects, and process forms.
🔹 Following best practices ensures smooth AI model performance.
🔹 AI-powered apps can revolutionize business processes with automation and intelligence.
By following these steps, you can successfully integrate AI Builder into your Power Apps and enhance your business applications with AI!
Would you like help with a specific AI Builder model for your Power Apps? 😊