Grouped by the problem each tool solves, not an alphabetical list. Bold marks the one to start with if you're choosing between options. Every tool links to its own docs or site, so you can go straight to the source.

Query and warehouse

• DuckDB: an in-process SQL engine, genuinely free forever, with no server to set up. The best first tool for practising SQL on real-sized data locally. Docs
• BigQuery sandbox: a free GCP warehouse tier, no credit card required
• Snowflake: a 30-day free trial (not perpetual, so good for a focused project rather than ongoing practice). Docs

Orchestration

• Apache Airflow (OSS, self-hosted or via Astronomer's free tier): the largest community and the most job-relevant. Docs
• Dagster (OSS): an asset-based model, good for dbt-centric projects. Docs

Transformation

• dbt Core: free, open-source, the industry-standard transformation tool. Docs

Ingestion and EL

• Airbyte (OSS): a free alternative to Fivetran, with a huge connector library. Docs
• Meltano: a CLI-first alternative, good if you're already comfortable in a terminal. Docs

Reverse ETL

• Hightouch: the category leader for destination breadth, moving data from your warehouse back out into the tools your team actually uses (CRM, ad platforms, support tools). Worth knowing this category exists before you need it: most data flows one way into a warehouse, reverse ETL is what sends it back out to power actions, not just reports.

Data catalog and governance

• DataHub (OSS): the most widely adopted open-source metadata catalogue, built by LinkedIn. Search, lineage and documentation for every table in your stack.
• OpenMetadata (OSS): a newer, actively developed alternative with data quality built into the catalogue itself, not bolted on separately.

BI and visualisation

• Metabase (OSS): the easiest free BI tool to self-host and actually enjoy using. Docs