Question

A developer at a marketing agency wants to quickly test how a large language model responds to different prompts before writing any application code. They have an Azure AI Foundry project and need the fastest way to deploy a model and experiment with prompts interactively. What should they do first in the Azure AI Foundry portal?

Accepted Answer

Deploy the model from the model catalog, then use the chat playground in the portal to send prompts and view responses

Answer

Install the Foundry SDK locally and write a Python script that authenticates and calls the model endpoint

Answer

Configure an Azure Content Understanding analyzer to extract fields from sample documents

Answer

Create a multi-agent orchestration and connect external tools before sending any prompts