A developer has used the Azure AI Foundry portal playground to manually test prompts against a candidate model and is satisfied with the results. The next step is to call that model automatically every night as part of an unattended batch pipeline. What should the developer do?
Select an answer to reveal the explanation.
Short Explanation
The playground is great for a person sitting there trying out prompts and eyeballing the results, but it has no way to run itself while everyone is asleep, so it is the wrong tool once the goal shifts to an unattended nightly job. What actually works is wiring the pipeline's own code straight to the deployed model's endpoint, whether through the SDK or a plain REST call, so the job can fire off requests on its own schedule with nobody watching. Exporting a chat transcript from a manual session just captures what already happened; it is not something the pipeline can call to get fresh answers on tonight's data. And spinning up a brand-new deployment for every single run, then tearing it down afterward, adds a pile of setup overhead for no real benefit when one deployment can just sit there and answer calls night after night.
Full Explanation
The correct answer is C. The playground in Azure AI Foundry is designed for interactive, manual experimentation with prompts and models, not for unattended, repeatable execution. Once a model has been validated there, moving it into production means calling the deployed model's REST endpoint or an SDK from the pipeline's own code, so the batch job can run automatically every night without a person present. Option A is incorrect because the playground requires a human to enter prompts and read results, which is incompatible with an unattended nightly job. Option B is incorrect because a chat history export is a log of a manual session, not a callable interface the pipeline can invoke to generate new responses on new data each night. Option D is incorrect because creating and deleting a fresh deployment for every run adds unnecessary setup time, cost, and complexity when a single persistent deployment can simply be called repeatedly through its endpoint.