Google Cloud Machine Learning

AI Portfolio

Big Data and AI/ML Products

Ingestion and Process	Storage	Analytics	Machine Learning
Pub Sub	Cloud Storage	BigQuery	Vertex AI
Dataflow	Cloud SQL	Looker Studio
Dataproc	Cloud Spanner	Looker
Cloud Data Fusion	Cloud Bigtable
	Cloud Firestore

Options for Building ML Models in Google Cloud

• BigQuery ML
• Prebuilt APIs
• AutoML—no-code solution in Vertex AI
• Custom training—control over ML pipelines in Vertex AI

Comparison

	BigQuery ML	Prebuilt APIs	AutoML	Custom Training
Data types	Tabular	Tabular, images, video, text	Tabular, images, video, text	Tabular, images, video, text
Training data size	Medium – large	N/A	Small – medium	Medium – large
ML and coding expertise	Medium	Low	Low	High
Flexibility to tune hyperparameters	Medium	None	None	High
Time to train	Medium	None	Medium	Long

Choosing

Considerations	Best Choice
Familiar with SQL, and data in BigQuery	BigQuery ML
Little ML experience	Prebuilt APIs
Custom models with own data, minimal coding	AutoML
Full control of ML workflow	Custom Training