Question

A retail company is building a Power Apps canvas app for store associates to log customer feedback. During data entry, the app must analyze the free-text feedback and immediately display whether the sentiment is positive, negative, or neutral so associates can flag unhappy customers for follow-up before they leave the store. The solution must require no custom model training and integrate directly into the canvas app formula bar. What should the architect recommend?

Accepted Answer

Add the prebuilt AI Builder sentiment analysis model and call it from a canvas app formula bound to the feedback text input

Answer

Train a custom AI Builder category classification model on historical feedback and invoke it via a Power Automate flow

Answer

Provision a custom model in Microsoft Foundry and expose it to the app through a custom connector

Answer

Build a Copilot Studio agent with generative orchestration and embed it in the app as a chat control