Question

You fine-tuned a GPT model in Azure AI Foundry for a customer-support workload. The model performed well on offline evaluation, but leadership requires evidence that it behaves correctly on real production traffic before it replaces the current model, without exposing customers to the new model's responses. Which promotion approach lets you validate the fine-tuned model against live traffic while ensuring users continue to receive responses only from the current production model?

Accepted Answer

Deploy the fine-tuned model as a shadow (mirror) deployment that receives a copy of production traffic while the existing deployment continues to serve all user responses.

Answer

Deploy the fine-tuned model with a canary rollout routing 10% of live traffic to it and returning its responses to those users.

Answer

Replace the current deployment with the fine-tuned model and monitor quality metrics, rolling back if regressions appear.

Answer

Run the fine-tuned model against a static offline evaluation dataset a second time with more test rows before promoting it.