Create a system diagram in draw.io demonstrating this solution and how your data flows through your components $1.$ The pipeline must run daily at midnight. $2.$ The internal data must be retrieved from an Azure SQL database $($the operational database$).3.$ The external data is delivered by a third party to a Google Cloud Storage bucket daily. $4.$ The model needs to be retrained daily $$ the training dataset is approximately $256$GB in size. $5.$ You wish to store the a feature$-$store$/$intermediate data for each run and are required to retain $120$ days of these data for audit purposes. $6.$ Model binaries are to produced for each run and deployed to a live endpoint for use by a digital origination application. $7.$ Model binaries need to be retained for $120$ days. $8.$ The live endpoint must have zero downtime. $9.$ Code and model logic changes need to be auditable. $10.$ Data drift and prediction drift monitoring is required. $11.$ All data must remain securely with Azure.