You now have the shape of the language. Here's where to go next.
Get fluent on a real database
- Install Postgres locally (or use Supabase / Neon for free hosted)
- Load a real dataset — your own bank statements, a Kaggle competition, NYC taxi trips
- Build one report end-to-end: from raw data to a chart
The next layer up: dbt
dbt (Data Build Tool) is how teams write production-grade SQL. It adds version control, tests, lineage, and modular models. If you're going to do data work for a living, this is what you'll use.
Visualization on top of SQL
Metabase (open source, free), Looker, Tableau, Superset, Power BI. They all write SQL under the hood; the difference is how much they let you bypass writing it. Knowing SQL means you'll never be limited by their UI.
Adjacent skills
- Python + pandas — when SQL stops being the right tool (machine learning, complex stats, scraping)
- Excel power-features — pivot tables, XLOOKUP, Power Query — for stakeholder-facing work
- Statistics — A/B testing, regression, time-series — to interpret what your queries return
- Communication — turn numbers into recommendations; the bottleneck is rarely the query
SQL dialects you'll meet
- Postgres — the default. Most powerful open-source SQL.
- SQLite — embedded, simple, perfect for learning. Subset of standard SQL.
- BigQuery — Google's analytics warehouse. SQL with extensions for arrays and structs.
- Snowflake — enterprise warehouse. Standard SQL with cloud-warehouse semantics.
- MySQL — older, still common. Mostly compatible; quirks around dates and booleans.
- DuckDB — Postgres-grade SQL on a single file. Fantastic for local analytics.
Where to spend your next 100 hours
Pick a real dataset that matters to you. Spend 10 hours getting it loaded. Then 90 hours asking questions of it and answering them with SQL. You'll learn more from one real project than from any course — including this one.
Exercise
You have completed this SQL course. You want to deepen your skills toward a data-analyst or BI-engineer role at a Kenyan bank or fintech within 12 months. Design your 90-day learning plan: (1) Specific concrete projects to build. (2) Tools and adjacent skills to add. (3) How to demonstrate the work publicly. (4) Where you would expect to be at the end of 90 days.