In this weekend after my birthday, I want to share this small proof of concept, in which I wanted to integrate some of my passions ML. NET, Azure Function and Power Apps.
As a first step I have created a model from ML.NET with Model Builder selecting "Select Value prediction Scenario", I leave you the link of a publication where I will explain step by step. (ML.NET & Model Builder and (Bonus Track) ASP.NET App to consume the model. Enjoy it! 🤟 - DEV Community 👩 💻👨 💻 )
The next goal was to be able to make inferences and predictions to the model from the cloud, for that use Azure Function. I share the link of my GitHub repository (ppiova/STEMsalariesAzureFunctions)
All of you must do is publish Azure Functions to Azure. (If you wish in the comments, in a next post I make a step like doing it).
So far, I already have my Machine Learning model, in addition the function is published in Azure as an API, using Postman I can test and make predictions with my model.
And now to be able to call this API published in Azure Function I have created a Custom Connector that I will then add in the application created with Power Apps.
I share the Github link (ppiova/PowerAppsSTEMSalaries) where you will find the json file to create the Custom Connector, an xlsx file to import it in the creation of the SharePoint list where you will save the predictions made and finally our strawberry of the dessert, haha, the application developed in Power Apps.
That's all folks, if you have questions about how to carry out this process, you write to me through my social networks, or in the comments.
Soon in addition to publishing posts I will start a new personal project in sharing content from my YouTube channel. I leave the link Pablito Piova - YouTube, Subscribe! 😊
Top comments (0)