AI helps humans and machines work together to make the world better. It helps the human brain handle the dehumanizing workload or repetitive tasks but not everyone has access to it. If we think about AI, it is complex and difficult to learn. Other than that, not all businesses have resources, like Data Analyst or Developer. They only have business users that can analyze data on the ground. But, what if business users can build AI easily and infuse it with their application? With this, it will bring the businesses to the next level. In the journey of democratizing of AI in Microsoft, AI Builder has been introduced.
AI Builder is now in General Availability Release. It is a Point-and-Click AI Solutions. Power Platform helps us build an application using Power Apps, create an automation process using Power Automate, and create data visualization using Power BI. Power Platform can connect all our data using the 250+ data connectors with other business applications and if we infuse it with AI, this helps you bring intelligence in your business application. AI Builder is a low code AI capability leveraging the Azure AI technologies that help the business build without limits.
As of this writing, all the AI Model (Form Processing, Object Detection, and Prediction) in AI Builder are in General Availability. The same goes with the other Prebuilt AI Models which are Business Card Reader, Sentiment Analysis, Language Detection, Key Phase Extraction and Text Recognition. For Prebuilt AI Model, you don’t need to train and publish this model. You can readily use it in your Flow and your Apps.
In three simple steps, we can create and infuse AI in our application using Form-Processing MODEL in AI Builder
- 1. Create your Form Processing Model
In creating your model, you need to upload 5 documents in minimum which can be up to 500 pages and should not exceed to 50MB for PDF documents. Learn more about the requirements and limitations.
- 2. Train your Form Processing Model
After analyzing the documents, you need to select the fields that you want to extract in the documents. Once done selecting the fields, you need to train your model. Training the models means your models will be based from the information or the fields that you’ve selected.
3. Publish your Form Processing Model
Once the training is done, you need to Publish your model to use it in your Power Automate or Power Apps application. You can do a Quick Test before publishing your model.
Here is the snippet below on how I used those steps in building my AI model using Form Processing Model where I use it in my Power Automate to extract the data from Email to my SharePoint list.