AI Builder for Power Platform is a hands-on course in using AI Builder to read invoices, receipts and forms, analyse text, build prompts and train document processing models, then use them in Power Automate and Power Apps with human review. It suits Power Platform users and accounting, purchasing and admin teams who key in data from large volumes of documents.
Document work in many organisations still relies on people keying in data: invoices, receipts, request forms and customer emails that have to be read and sorted one by one. Many teams already use Power Automate and Power Apps but do not realise that AI Builder can read documents, extract data, classify text and summarise content out of the box, with no code and no need to build AI models themselves, or where to put human review so the results can be trusted.
This course teaches AI Builder systematically. Learners start with AI hub and prebuilt models for invoices, receipts, text recognition and text analysis, then build prompts that take text, documents or images and return JSON, with responsible AI practices. They go on to build a custom document processing model for their own documents, use every model in Power Automate and Power Apps, add human review based on confidence scores and plan Copilot Credits consumption. The course closes with a capstone that builds an automated document processing system. (2 days, 6 hours per day, 12 hours in total, Intermediate level.)
Course Details
This course is entirely about using AI Builder on Power Platform. It runs for 2 days, 6 hours per day (12 hours in total, 09:00-16:00), as lectures with labs and workshops on a sample company's documents. Intermediate level. It builds on Power Automate Cloud for Business and Microsoft Power Apps, which only introduce AI Builder briefly, and covers prebuilt models for invoices, receipts and text, prompts that take documents or images, document processing models, custom text models, use in Power Automate and Power Apps, human review and credit planning. Learners use their own or their organisation's premium rights; the institute does not provide accounts, licences, credits or environments. From 1 November 2026, AI Builder usage under new licences draws on Copilot Credits instead of AI Builder credits.
Section 1: Getting to Know AI Builder and AI hub
Section 2: Lab: Invoices and Receipts
Section 3: Lab: Reading Text from Images and Other Documents
Section 4: Lab: Analysing Text
Section 5: Lab: Building Prompts in Power Platform
Section 6: Reliable Prompts and Responsible AI
Section 7: Lab: Document Processing Models
Section 8: Lab: Custom Text Models
Section 9: Lab: AI Builder in Power Automate
Section 10: Lab: Human Review and Exceptions
Section 11: Lab: AI Builder in Power Apps
Section 12: Workshop: Automated Document System Capstone
No public rounds are open right now. Join the waiting list and we will contact you first when the next round opens, or ask us on LINE. Or call 02-570-8449 or 088-807-9770
This course builds on Power Automate Cloud for Business and Microsoft Power Apps, which each give AI Builder a single topic. This course spends both days on AI Builder: prebuilt models, prompts, document processing models and custom text models, through to use in flows and apps, human review and credit planning.
The institute does not provide accounts, licences, credits or environments. You need your own or your organisation's Power Automate Premium or Power Apps Premium rights, in an environment with AI Builder enabled and enough AI Builder credits or Copilot Credits. Check with your Power Platform admins before the training day.
Microsoft stopped selling AI Builder capacity add-ons on 1 November 2025, and from 1 November 2026 newly purchased premium licences no longer include seeded AI Builder credits. Existing credits remain usable until the end of the contract, after which usage draws on Copilot Credits. AI Builder features themselves keep working. The class explains how to estimate and track credit usage.
It depends on the model. Prompts support Thai, while each prebuilt model and document processing supports different languages. The class shows how to check language support and test with Thai documents, and how to use a prompt instead where a prebuilt model does not yet support Thai. Real results depend on document quality and layout.
No. Every lab uses the low-code Power Platform interface, but you should be able to build a basic cloud flow or have built a canvas app, so you can use the models in flows and apps comfortably.
The main labs use sample documents. If you want to try real documents, use only those you are permitted to use, remove personal or confidential data, and work in your organisation's environment in line with its data policies and PDPA.
Many organisations still run work on paper, on Excel files passed through several hands, or on forms whose data has to be retyped into another system. Writing a new program in a programming language takes time and depends on a development team. Microsoft…
A lot of office work is still repeated by hand every day: copying data from emails into Excel, sending forms to a manager for approval, chasing documents that have not arrived or letting the team know in Teams when a new item comes in. Organisations that use…
Many organisations already use Microsoft 365 Copilot for everyday work. The next step is building their own AI agents that know the organisation's data, answer questions from internal policies and documents, and take action on people's behalf, such as opening…
