BigQuery — Serverless analytics warehouse
A serverless SQL data warehouse. Load billions of rows, run analytical queries in seconds, pay per byte scanned. Not a transactional database.
is a giant SQL engine for analytics. You throw billions of rows at it (web events, application logs, sales records, IoT data) and ask questions in SQL — 'what's the weekly revenue by country for the last 2 years' — and it answers in seconds.
It's serverless: no cluster to size, no instances to manage, no indexes to tune. You run SQL, Google's backend spreads the work over thousands of machines, returns the answer. You pay for the bytes the query scanned (or, alternatively, for reserved capacity).
BigQuery is a data warehouse, not a database for your app. Don't put your users table there to read it on every login — a query takes hundreds of milliseconds minimum and isn't made for single-row lookups. Use or Firestore for that.
Typical uses: BI dashboards (connect Looker or Tableau), ad-hoc analytics, machine learning (BigQuery ML lets you train models with SQL), real-time log analysis (streaming inserts from or Cloud Logging exports).
Comparison: BigQuery = Snowflake = AWS Redshift Serverless = Azure Synapse. BigQuery stands out for being fully serverless from day one, with deep integration across the GCP ecosystem and a generous free tier (1 TB scanned/month free).
Grounded on https://cloud.google.com/bigquery/docs/introduction
Next up
Vertex AI — Google's ML platform
A unified platform to train, tune, deploy, and call ML models — including Google's Gemini family via the Gemini API. Covers the full ML lifecycle.