The correct approach is to enable Azure Synapse Link on the Azure Cosmos DB account and configure the SQL pool to access the replicated data. This allows the SQL pool to query the data that has been replicated from Azure Cosmos DB. Option A is incorrect because creating an external table in the SQL pool is not sufficient to access the replicated data. Option B is incorrect because Azure Data Factory is used for moving data between different storage services, not for replicating data. Option C is incorrect because Azure Databricks is used for processing data, not for replicating data.