Understanding Different AI Builder Models in Power Automate

Introduction

In this article, we will understand all AI Builder models, what they do, and where they are used, explained in simple and easy way.

What is AI Builder?

AI Builder in Power Automate is a Microsoft tool that lets you use Artificial Intelligence (AI) in your flows without writing any code.

With AI Builder, you can:

Read documents
Understand text
Detect objects in images
Predict outcomes
Process invoices
Extract structured data

Types of AI Builder Models

AI Builder models are mainly divided into Prebuilt models and Custom models.

Prebuilt AI Builder Models

Prebuilt models are ready to use. You don’t need to train them. Microsoft has already trained these models for common business scenarios.

Invoice Processing Model
What is it?
Invoice Processing is a prebuilt AI model that automatically reads invoice files and extracts important information.
What kind of files does it support?
- PDF invoices
- Scanned invoice images
- Photos of invoices
What data does it extract?
- Invoice number
- Invoice date
- Vendor name
- Vendor address
- Customer name
- Subtotal
- Tax amount
- Total amount
- Due date
- Line items (product name, quantity, price)
Why is it useful?
- Removes manual invoice data entry
- Reduces human errors
- Saves time for finance teams
Where is it commonly used?
- Accounts payable systems
- Finance automation
- Vendor billing solutions
Receipt Processing Model
What is it?
Receipt Processing extracts data from receipts such as restaurant bills or shopping receipts.
What data does it extract?
- Merchant name
- Transaction date
- Total amount
- Tax amount
- Payment method
Use cases
- Expense management
- Employee reimbursement systems
- Travel and food expense tracking
Business Card Reader Model
What is it?
This model reads business cards and extracts contact details.
Extracted details
- Person name
- Company name
- Email address
- Phone number
- Job title
- Address
Use cases
- CRM systems
- Lead management
- Contact creation in Dataverse
Identity Document Reader
What is it?
This model extracts data from identity documents.
Supported documents
- Passport
- Driving license
- National ID cards
Extracted data
- Full name
- Date of birth
- Document number
- Expiry date
- Country
Use cases
- Employee onboarding
- Customer verification
- Compliance processes
Text Recognition (OCR) Model
What is it?
This model reads text from images or scanned documents.
What can it do?
- Extract printed text
- Extract handwritten text (limited support)
Use cases
- Reading scanned documents
- Converting images to text
- Searching text in documents
Language Detection Model
What is it?
Detects the language of given text automatically.
Supported languages
- English
- Hindi
- French
- Spanish
- Many more
Use cases
- Multi-language apps
- Customer support automation
- Global business solutions
Sentiment Analysis Model
What is it?
This model understands whether text is positive, negative, or neutral.
Examples
- Customer feedback
- Reviews
- Survey responses
Use cases
- Customer satisfaction analysis
- Product feedback analysis
- Support ticket analysis
Key Phrase Extraction Model
What is it?
Extracts important keywords from text.
Example
Input text:
“This product delivery was late and support was slow.”
Output:
- Delivery
- Support
- Product
Use cases
- Feedback analysis
- Document summarization
- Search optimization
Category Classification Model
What is it?
Classifies text into predefined categories.
Example categories
- Complaint
- Request
- Feedback
- Inquiry
Use cases
- Email categorization
- Support ticket routing
- Helpdesk automation

Custom AI Builder Models

Custom models are created when prebuilt models don’t meet specific needs.

Custom Document Processing Model

What is it?
Used to extract data from documents that follow a custom format.
Examples
- Purchase orders
- Contracts
- Custom reports
- Industry-specific documents
Use Case
Allows Power Automate to understand documents unique to your organization.
Prediction Model
What is it?
Predicts outcomes using historical data.
Examples
- Predict late payments
- Forecast sales outcomes
- Identify risk cases
Use Case
Helps decision-making by adding intelligence to automated processes.
Object Detection Model
What is it?
Detects and identifies objects in images.
Examples
- Identify damaged products
- Count items
- Detect equipment types
Use Case
Useful in inventory management and quality checks.

Conclusion

AI Builder models make Power Automate smarter by understanding text, documents, and data, not just moving files or sending emails. Using these models, anyone can create smarter and more efficient automations easily.