Topic hub

BigQuery for Analytics Engineers

Google's serverless warehouse, used as the backbone of our capstone — partitioning, clustering, BigQuery SQL patterns, and dbt on BigQuery.

BigQuery is Google's serverless data warehouse — separation of storage and compute, no infrastructure to manage, and a free tier generous enough to learn on without a credit card surprise. It's the warehouse we recommend for first-time analytics engineers and the one our capstone is built on.

The articles below cover the BigQuery-specific patterns that come up in real work: writing efficient SQL, controlling cost with partitioning and clustering, ingesting columnar data, and integrating dbt against BigQuery. Pair them with the SQL hub for the language and the dbt hub for the framework.

What you'll learn

By the end of this path you can…

  • Spin up a BigQuery project and run your first GoogleSQL query
  • Use partitioning and clustering to control cost on large tables
  • Write idiomatic BigQuery SQL (array/struct, ML.PREDICT, etc.)
  • Connect BigQuery to dbt Cloud and ship a real project
  • Reason about BigQuery billing before queries run, not after
The learning path

From beginner to job-ready.

  1. 01 · Foundations

    Projects, datasets, tables — the BigQuery hierarchy and what slots are.

  2. 02 · SQL on BigQuery

    Best practices for GoogleSQL, including array/struct and timestamp patterns.

  3. 03 · Cost control

    Partitioning, clustering, materialized views, and reading the query estimator.

  4. 04 · Ingest

    Loading CSVs, JSONL, and Parquet; external tables; scheduled queries.

  5. 05 · dbt + BigQuery

    Wiring dbt Cloud to BigQuery, the capstone build, and CI patterns.

In the course

Analytics Engineering Capstone Project

27 lessons in this module

Common questions

Common questions about this topic.

Is BigQuery free to learn on?

Mostly yes. The sandbox tier gives you 10 GB of active storage and 1 TB of query processing per month at no cost. For the capstone in this course, you'll stay inside that envelope unless you deliberately scale beyond it.

BigQuery or Snowflake for a portfolio project?

Either signals well. We use BigQuery in the capstone for the easier on-ramp (no credit card for the sandbox, less ops). If a target employer is a Snowflake shop, doing the project on Snowflake is equally valid — the dbt code is nearly identical.

How do partitioning and clustering actually help?

Partitioning lets BigQuery prune which storage blocks to scan, slashing the bytes-billed for date-filtered queries. Clustering co-locates rows with similar values, making filter and join predicates cheaper. Combine them on large fact tables and watch your bill drop.

Does dbt run on BigQuery?

Yes — first-class. The dbt-bigquery adapter is mature, and dbt Cloud has a native BigQuery connector. The capstone walks through the full wiring.

Start practicing this topic.

Graded exercises with hints, worked solutions, and a GPT tutor. Free to start, no credit card.