Skip to content

Quick Start

Get hands-on with BigQuery ETL Skills through practical examples.

Your First Query

Let's create a simple aggregation query from scratch:

Write a query that counts daily active users from clients_daily for Firefox Desktop

Claude will:

  1. ✅ Use bigquery-etl-core to understand conventions
  2. ✅ Use query-writer to create the SQL
  3. ✅ Use metadata-manager to generate schema and metadata files
  4. ✅ Optionally use sql-test-generator to create unit tests

Common Tasks

Create Unit Tests

Create unit tests for telemetry_derived/events_daily_v1

Update an Existing Query

Add a new field to telemetry_derived/clients_daily_v6
that calculates the ratio of searches with ads

Gather Requirements

I need to create a table tracking campaign click-through rates.
Help me gather requirements.

Add Monitoring

Add Bigeye monitoring for telemetry_derived/clients_daily_v6

How It Works

Automatic Skill Selection

Claude automatically selects the right skills based on your request. You don't need to explicitly invoke them—just describe what you want naturally!

When you make a request, Claude:

  1. Analyzes your intent
  2. Selects appropriate skills
  3. Executes the workflow
  4. Iterates based on your feedback

Tips for Success

Be Specific

  • ✅ "Create a daily aggregation query for active users"
  • ❌ "Help with a query"

Include Paths

Mention specific tables: telemetry_derived/clients_daily_v6

Work in Context

Launch Claude from your bigquery-etl directory for file operations