A legal team wants their Foundry-deployed chat model to answer questions using the content of their internal contract repository stored in Azure Blob Storage, without retraining or fine-tuning the underlying model. Which Foundry feature should they configure to link that storage account so retrieved contract text augments the model's responses at query time?

The correct answer is A. Configuring a data connection lets Foundry index and retrieve relevant contract text from the linked storage account at query time, grounding the model's responses in that content through retrieval-augmented generation, all without altering the model's underlying weights. Option B is incorrect because fine-tuning retrains the model on the repository's content, which directly contradicts the stated requirement to avoid retraining, and it is also poorly suited to a contract repository that changes over time since the model would need retraining after every update. Option C is incorrect because the maximum output token limit only controls how long a generated response can be; it has no effect on whether the model can access or retrieve external contract documents. Option D is incorrect because matching the deployment region to the storage account's region may reduce network latency, but proximity alone does not connect the storage account as a retrievable data source; the model still needs an explicit data connection to know the content exists and pull from it. A data connection is the mechanism that ties external content to the model's responses at query time.