This document includes the best practices and guidelines for BigQuerywhen running generative AI workloads on Google Cloud. UseBigQuerywith Vertex AI to store data. Using BigQuery withVertex AI can significantly enhance your ML workflow because you can simplifydata access, enable scalable analysis, and use its ML capabilities.

Consider the following use cases for BigQuery with Vertex AI:

• Seamless integration: BigQuery and Vertex AIare tightly integrated, letting you access and analyze your data directlywithin the Vertex AI platform. This integration eliminates theneed for data movement, streamlines your ML workflow, and reduces friction.
• Scalable data analysis: BigQuery offers a petabyte-scaledata warehouse,letting you analyze massive datasets without worrying about infrastructurelimitations. This scalability is critical for training and deploying MLmodels that require vast amounts of data.
• SQL-based ML: BigQuery ML lets you use familiar SQL commands totrainand deploy models directly within BigQuery. This feature lets data analystsand SQL practitioners use ML capabilities without requiring advanced codingskills.
• Online and batch predictions: BigQuery ML supports onlineand batchpredictions. You can run real-time predictions on individual rows orgenerate predictions for large datasets in batch mode. This flexibilitypermits diverse use cases with varying latency requirements.
• Reduced data movement: With BigQuery ML, you don't need tomove yourdata to separate storage or compute resources for model training anddeployment. This reduced movement simplifies your workflow, reduces latency,and minimizes cost associated with data transfer.
• Model monitoring: Vertex AI provides comprehensive modelmonitoringcapabilities, letting you track the performance, fairness, andexplainability of your BigQuery ML models. Model monitoring helps you ensurethat your models are performing as expected and address potential issues.
• Pretrained models: Vertex AI offers access to pretrainedmodels,including those for natural language processing and computer vision. You canuse these models within BigQuery to enhance your analysis and extract deeperinsights from your data.
• Cost-effective solution: BigQuery ML offers acost-effective, flexibleway to train and deploy ML models. You only pay for the resources you use,making it an affordable option for organizations of all sizes.
• Advanced analytics capabilities: BigQuery provides toolsfor advancedanalytics, including geospatial analysis and forecasting. These tools letyou combine ML with other analytical techniques for deeper data explorationand richer insights.
• Enhanced collaboration: By using BigQuery withVertex AI, datascientists, ML engineers, and analysts can collaborate seamlessly on MLprojects. This collaboration helps create a more integrated and efficientapproach to tackling complex data problems.

Required BigQuery controls

The following controls are strongly recommended when usingBigQuery.

Ensure BigQuery datasets aren't publicly readable or set to allAuthenticatedUsers

Ensure BigQuery tables aren't publicly readable or set to allAuthenticatedUsers

Optional BigQuery controls

These controls are optional. Consider enforcing them when they apply to your specific use cases.

Encrypt individual values in a BigQuery table

Use authorized views for BigQuery datasets

Use BigQuery column-level security

Use BigQuery row-level security

Use BigQuery resource charts

What's next

• ReviewCloud Billingcontrols.

• See moreGoogle Cloud security best practices and guidelines for generativeAI workloads.